Irad Ben-Gal

It has been claimed that many security breaches are often caused by vulnerable (naïve) employees within the organization [Ponemon Institute LLC 2015a]. Thus, the weakest link in security is often not the technology itself but rather the people who use it [Schneier 2003]. In this paper, we propose a machine learning scheme for detecting risky webpages and risky browsing behavior, performed by naïve users in the organization. The scheme analyzes the interaction between two modules: one represents… רוצה לראות יותר

It has been claimed that many security breaches are often caused by vulnerable (naïve) employees within the organization [Ponemon Institute LLC 2015a]. Thus, the weakest link in security is often not the technology itself but rather the people who use it [Schneier 2003]. In this paper, we propose a machine learning scheme for detecting risky webpages and risky browsing behavior, performed by naïve users in the organization. The scheme analyzes the interaction between two modules: one represents naïve users, while the other represents risky webpages. It implements a feedback loop between these modules such that if a webpage is exposed to a lot of traffic from risky users, its "risk score" increases, while in a similar manner, as the user is exposed to risky webpages (with a high "risk score"), his own "risk score" increases. The proposed scheme is tested on a real-world dataset of HTTP logs provided by a large American toolbar company. The results suggest that a feedback learning process involving webpages and users can improve the scoring accuracy and lead to the detection of unknown malicious webpages. רוצה לראות פחות

This work proposes a method to improve the QoS provided to internet users by website servers. In particular, the goal is to minimize the expected number of browsing steps (clicks), also known as the expected path length, which are required to reach a website page by a community of users. We use Markov chain modeling to represent the transition probabilities from one webpage to another, and the first passage from page to page that can be retrieved from web server logs. The proposed
method… רוצה לראות יותר

This work proposes a method to improve the QoS provided to internet users by website servers. In particular, the goal is to minimize the expected number of browsing steps (clicks), also known as the expected path length, which are required to reach a website page by a community of users. We use Markov chain modeling to represent the transition probabilities from one webpage to another, and the first passage from page to page that can be retrieved from web server logs. The proposed
method deletes links among webpages to minimize the expected path length of the website. Three different methods for achieving this goal are examined: (i) a greedy deletion heuristic; (ii) an approximated branch and bound algorithm; and (iii) a cross-entropy metaheuristic. Numerical studies show that the proposed greedy heuristic results in the optimal solution in more than 60% of the tested cases, while in almost 90% of the cases, the obtained solution is within 10% of the optimal solution. Both the approximated branch and bound and the cross-entropy methods achieved optimality in more than 80% of the tested cases; however, this came with a much higher computational cost. רוצה לראות פחות

In recent years, with the emergence of big data and online Internet applications, the ability to classify huge amounts of objects in a short time has become extremely important. Such a challenge can be achieved by constructing decision trees (DTs) with a low expected number of tests (ENT). We address this challenge by proposing the ‘save favorable general optimal testing algorithm’ (SFGOTA)
that guarantees, unlike conventional look-ahead DT algorithms, the construction of DTs with monotonic… רוצה לראות יותר

In recent years, with the emergence of big data and online Internet applications, the ability to classify huge amounts of objects in a short time has become extremely important. Such a challenge can be achieved by constructing decision trees (DTs) with a low expected number of tests (ENT). We address this challenge by proposing the ‘save favorable general optimal testing algorithm’ (SFGOTA)
that guarantees, unlike conventional look-ahead DT algorithms, the construction of DTs with monotonic non-increasing ENT. The proposed algorithm has a lower complexity in comparison to conventional look-ahead algorithms. It can utilize parallel processing to reduce the execution time when needed. Several numerical studies exemplify how the proposed SF-GOTA generates
efficient DTs faster than standard look-ahead algorithms, while converging to a DT with a minimum ENT. רוצה לראות פחות

An online group testing method to search for a hidden object in a discrete search space is proposed. A relevant example is a search after a nonconforming unit in a batch, while many other applications can be related. A probability mass function is defined over the search space to represent the probability of an object (e.g., a nonconforming unit) to be located at some point or subspace. The suggested
method follows a stochastic local search procedure and can be viewed as a generalization of… רוצה לראות יותר

An online group testing method to search for a hidden object in a discrete search space is proposed. A relevant example is a search after a nonconforming unit in a batch, while many other applications can be related. A probability mass function is defined over the search space to represent the probability of an object (e.g., a nonconforming unit) to be located at some point or subspace. The suggested
method follows a stochastic local search procedure and can be viewed as a generalization of the Learning Real-Time A∗ (LRTA∗) search algorithm, while using informational distance measures over the searched space. It is proved that the proposed Informational LRTA∗ (ILRTA∗) algorithm converges and always terminates. Moreover, it is shown that under relevant assumptions, the proposed algorithm generalizes known optimal information-theoretic search procedures, such as the offline Huffman search or the generalized
optimum testing algorithm. However, the ILRTA∗ can be applied to new situations, such as a search with side information or an online search where the probability distribution changes. The obtained results can help to bridge the gap between different search procedures that are related to quality control, artificial intelligence, and information theory. רוצה לראות פחות

The construction of efficient decision and classification trees is a fundamental task in Big Data
analytics which is known to be NP-hard. Accordingly, many greedy heuristics were suggested for the construction of decision-trees, but were found to result in local-optimum solutions. In this work we present the dual information distance (DID) method for efficient construction of decision trees that is computationally attractive, yet relatively robust to noise. The DID heuristic selects features… רוצה לראות יותר

The construction of efficient decision and classification trees is a fundamental task in Big Data
analytics which is known to be NP-hard. Accordingly, many greedy heuristics were suggested for the construction of decision-trees, but were found to result in local-optimum solutions. In this work we present the dual information distance (DID) method for efficient construction of decision trees that is computationally attractive, yet relatively robust to noise. The DID heuristic selects features by considering both their immediate contribution to the classification, as well as their future potential effects. It represents the construction of classification trees by finding the shortest paths over a graph of partitions
that are defined by the selected features. The DID method takes into account both the orthogonality between the selected partitions, as well as the reduction of uncertainty on the class partition given the selected attributes. We show that the DID method often outperforms popular classifiers, in terms of average depth and classification accuracy. רוצה לראות פחות

In the paper, a complete minimal system of fuzzy logic operators that implements the operations over qubits is suggested. The model is based on a parametric system of the fuzzy not and phase operators and represents operators of quantum information theory acting on real amplitudes. An example of the system application is provided by its implementation to a mobile robot control, and its correspondence with the previously suggested models is illustrated by numerical simulations.

Condition-based maintenance (CBM) is increasingly applied to operational systems to reduce lifecycle costs. Predicting the performance of various CBM policies is a challenging task addressed in this work. We suggest a CBM framework that is based on system simulations and a targeted Bayesian network model. Simulations explore the robustness of various CBM policies under different scenarios. The Bayesian network, which is learned from the simulation data, is then used as an explanatory compact… רוצה לראות יותר

Condition-based maintenance (CBM) is increasingly applied to operational systems to reduce lifecycle costs. Predicting the performance of various CBM policies is a challenging task addressed in this work. We suggest a CBM framework that is based on system simulations and a targeted Bayesian network model. Simulations explore the robustness of various CBM policies under different scenarios. The Bayesian network, which is learned from the simulation data, is then used as an explanatory compact metamodel for failure prediction. The framework is demonstrated through a study of an operator of a freight rail fleet. This study demonstrates a significant profit improvement compared to other methods. רוצה לראות פחות

This article investigates methods for reallocation of service capacities in open Jackson networks in order to minimize either a system’smean total work-in-process or its response time. The focus is mainly on a method called node generation, by which capacity can be transferred from a node in echelon j to a newly generated node in echelon j + 1. The proposed procedure is compared with
the more conventional capacity redistribution method, by which capacity can be transferred from any node in… רוצה לראות יותר

This article investigates methods for reallocation of service capacities in open Jackson networks in order to minimize either a system’smean total work-in-process or its response time. The focus is mainly on a method called node generation, by which capacity can be transferred from a node in echelon j to a newly generated node in echelon j + 1. The proposed procedure is compared with
the more conventional capacity redistribution method, by which capacity can be transferred from any node in echelon j to existing
successor nodes in echelon j + 1. Formulation of each method as a mathematical programming problem reveals the structure of the optimal solution for both problems. The motivation for considering these approaches stems from real-life settings, in particular, from a production line or supply chains where the two types of capacity reallocation are applied. Heuristic methods are developed to solve relatively large networks in tractable time. Numerical results and analyses are presented. רוצה לראות פחות

This article suggests a new Statistical Process Control (SPC) approach for data-rich environments. The proposed approach is based on the theory of fractal geometry. In particular, a monitoring scheme is developed that is based on fractal representation of the monitored data at each stage to account for online changes in monitored processes. The proposed fractal-SPC enables a dynamic inspection of non-linear and state-dependent processes with a discrete and finite state space. It is aimed for… רוצה לראות יותר

This article suggests a new Statistical Process Control (SPC) approach for data-rich environments. The proposed approach is based on the theory of fractal geometry. In particular, a monitoring scheme is developed that is based on fractal representation of the monitored data at each stage to account for online changes in monitored processes. The proposed fractal-SPC enables a dynamic inspection of non-linear and state-dependent processes with a discrete and finite state space. It is aimed for use with both univariate and multivariate data. The SPC is accomplished by applying an iterated function system to represent a process as a fractal and exploiting the fractal dimension as an important monitoring attribute. It is shown that data patterns can be transformed into representing fractals in a manner that preserves their reference (in control) correlations and dependencies. The fractal statistics can then be used for anomaly detection, pattern analysis, and root cause analysis. Numerical examples and comparisons to conventional SPC methods are given. רוצה לראות פחות

Peer-to-peer (p2p) networks are used by millions for searching and downloading
content. Recently, clustering algorithms were shown to be useful for helping users find
content in large networks. Yet, many of these algorithms overlook the fact that p2p networks
follow graph models with a power-law node degree distribution. This paper studies
the obtained clusters when applying clustering algorithms on power-law graphs and their
applicability for finding content. Driven by the… רוצה לראות יותר

Peer-to-peer (p2p) networks are used by millions for searching and downloading
content. Recently, clustering algorithms were shown to be useful for helping users find
content in large networks. Yet, many of these algorithms overlook the fact that p2p networks
follow graph models with a power-law node degree distribution. This paper studies
the obtained clusters when applying clustering algorithms on power-law graphs and their
applicability for finding content. Driven by the observed deficiencies, a simple yet efficient
clustering algorithm is proposed, which targets a relaxed optimization of a minimal distance
distribution of each cluster with a size balancing scheme. A comparative analysis using a
song-similarity graph collected from 1.2 million Gnutella users reveals that commonly used
efficiency measures often overlook search and recommendation applicability issues and provide
the wrong impression that the resulting clusters are well suited for these tasks. We show
that the proposed algorithm performs well on various measures that are well suited for the
domain. רוצה לראות פחות

We consider an Ishida and Korf Moving Target Search (MTS) algorithm with informational distance measures. Similarly to the previously defined Informational Learning Real-Time A* algorithm, the suggested algorithm acts on the set of partitions of the sample space, on which the probability mass function is defined. The information-based Rokhlin metric and its lower bound – Ornstein metric, give the necessary distance measures. We prove that similarly to the Ishida
and Korf MTS algorithm, the… רוצה לראות יותר

We consider an Ishida and Korf Moving Target Search (MTS) algorithm with informational distance measures. Similarly to the previously defined Informational Learning Real-Time A* algorithm, the suggested algorithm acts on the set of partitions of the sample space, on which the probability mass function is defined. The information-based Rokhlin metric and its lower bound – Ornstein metric, give the necessary distance measures. We prove that similarly to the Ishida
and Korf MTS algorithm, the proposed Informational MTS (IMTS) algorithm always terminates and finds the target. The comparison of the IMTS algorithm with known models shows that it outperforms known Markov decision process model of search with probabilistic and informational decision criteria. These findings help to construct a unified framework of search after both static and moving targets, and to bridge the gap between different search procedures that are related to
both artificial intelligence and information theory. רוצה לראות פחות

Condition-based maintenance (CBM) is increasingly applied to operational systems to reduce lifecycle costs. Predicting the performance of various CBM policies is a challenging task addressed in this work.
We suggest a CBM framework that is based on system simulations and a targeted Bayesian network model. Simulations explore the robustness of various
CBM policies under different scenarios. The Bayesian network, which is learned from the simulation data, is then used as an explanatory… רוצה לראות יותר

Condition-based maintenance (CBM) is increasingly applied to operational systems to reduce lifecycle costs. Predicting the performance of various CBM policies is a challenging task addressed in this work.
We suggest a CBM framework that is based on system simulations and a targeted Bayesian network model. Simulations explore the robustness of various
CBM policies under different scenarios. The Bayesian network, which is learned from the simulation data, is then used as an explanatory compact metamodel for failure prediction. The framework is demonstrated through a study of an operator of a freight rail fleet. This study demonstrates a significant profit improvement compared to other methods רוצה לראות פחות

We present a new Bayesian network modeling that learns the behavior of an unknown system
from real data and can be used for reliability engineering and optimization processes in industrial systems. The suggested approach relies on quantitative criteria for addressing the trade-off between the complexity of a learned model and its prediction accuracy. These criteria are based on measures from Information Theory as they predetermine both the accuracy as well as the complexity of the model. We… רוצה לראות יותר

We present a new Bayesian network modeling that learns the behavior of an unknown system
from real data and can be used for reliability engineering and optimization processes in industrial systems. The suggested approach relies on quantitative criteria for addressing the trade-off between the complexity of a learned model and its prediction accuracy. These criteria are based on measures from Information Theory as they predetermine both the accuracy as well as the complexity of the model. We illustrate the proposed method by a classical example of system reliability engineering. Using computer experiments, we show how in a targeted Bayesian network learning, a tremendous reduction in the model complexity can be accomplished, while maintaining most of the essential information for optimizing the system. רוצה לראות פחות

The weak form of the Efficient Market Hypothesis (EMH) states that the current market price
fully reflects the information of past prices and rules out predictions based on price data alone. In
an efficient market, consistent prediction of the next outcome of a financial time series is problematic
because there are no reoccurring patterns that can be used for a reliable prediction.
This research offers an alternative test of the weak form of the EMH. It uses a universal… רוצה לראות יותר

The weak form of the Efficient Market Hypothesis (EMH) states that the current market price
fully reflects the information of past prices and rules out predictions based on price data alone. In
an efficient market, consistent prediction of the next outcome of a financial time series is problematic
because there are no reoccurring patterns that can be used for a reliable prediction.
This research offers an alternative test of the weak form of the EMH. It uses a universal prediction
algorithm based on the Variable Order Markov tree model to identify re-occurring patterns
in the data, constructs explanatory models, and predicts the next time-series outcome. Based on
these predictions, it rejects the EMH for certain stock markets while accepting it for other markets.
The weak form of the EMH is tested for four international stock exchanges: the German DAX
index; the American Dow-Jones30 index; the Austrian ATX index and the Danish KFX index.
The universal prediction algorithm is used with sliding windows of 50, 75, and 100 consecutive
daily returns for periods of up to 12 trading years. Statistically significant predictions are detected
for 17% to 81% of the ATX, KFX and DJ30 stock series for about 3% to 30% of the trading days.
A summary prediction analysis indicates that for a confidence level of 99% the more volatile German
(DAX) and American (DJ30) markets are indeed efficient. The algorithm detects periods of
potential market inefficiency in the ATX and KFX markets that may be exploited for obtaining
excess returns.
רוצה לראות פחות

Purpose
The purpose of this study is to develop a model for physician staffing requirements
that could be generally applied to any hospital department, taking into account factors such
as occupancy level, professional absences, allowances, physician work duties and
patient service levels.
Design/methodology/approach
The ability to generalize the model was tested via its implementation to five hospital
departments considered to represent a cross-section of all hospital… רוצה לראות יותר

Purpose
The purpose of this study is to develop a model for physician staffing requirements
that could be generally applied to any hospital department, taking into account factors such
as occupancy level, professional absences, allowances, physician work duties and
patient service levels.
Design/methodology/approach
The ability to generalize the model was tested via its implementation to five hospital
departments considered to represent a cross-section of all hospital requirements: Internal
Medicine, Surgery, Orthopedics Children's (Pediatrics), and Gynecology. The work is
based on a combination of a survey, work-sampling and direct time study, conducted
by professional observers with a high degree of frequency and over a relatively long
period of time.
Findings
The model that we developed is based on the concept of "required work capacity" of
physicians. The model makes it possible to account for such factors as yearly
capacity, level of desired service, increments for work conditions, roster duties and
necessary absences. רוצה לראות פחות

In this paper we explore the Vs-optimality criterion that was proposed in Ginsburg and Ben-Gal (IIE Trans. 2006; 38:445–461) as a new design-of-experiment (DOE) alphabetic optimality criterion. The Vs-optimality criterion seeks to minimize the variance of the optimal solution of an empirically fitted model. We show that the Vs-optimality citerion is well related to known alphabetic DOE criteria. However, it focuses on an ‘optimization-seeking’ experimental approach rather than an… רוצה לראות יותר

In this paper we explore the Vs-optimality criterion that was proposed in Ginsburg and Ben-Gal (IIE Trans. 2006; 38:445–461) as a new design-of-experiment (DOE) alphabetic optimality criterion. The Vs-optimality criterion seeks to minimize the variance of the optimal solution of an empirically fitted model. We show that the Vs-optimality citerion is well related to known alphabetic DOE criteria. However, it focuses on an ‘optimization-seeking’ experimental approach rather than an ‘information-seeking’ approach, which is often adopted by traditional optimality criteria. We illustrate the differences between these two approaches by a detailed example of a robust design of an RL-circuit. רוצה לראות פחות

Automotive assembly lines are often characterized by robots’ failures that may result in
stoppages of the lines and manual backup of tasks. The phenomena tend to impair
throughput rate and products’ quality. This paper presents a backup strategy in which
working robots perform tasks of failed robots. The proposed Mixed-Integer LinearProgramming
based approach minimizes the throughput loss by utilizing the robots’
redundancy in the system. Two algorithms are developed to comply… רוצה לראות יותר

Automotive assembly lines are often characterized by robots’ failures that may result in
stoppages of the lines and manual backup of tasks. The phenomena tend to impair
throughput rate and products’ quality. This paper presents a backup strategy in which
working robots perform tasks of failed robots. The proposed Mixed-Integer LinearProgramming
based approach minimizes the throughput loss by utilizing the robots’
redundancy in the system. Two algorithms are developed to comply with stochastic
conditions of a real-world environment. The performance of these algorithms is
compared with several heuristics, and the downstream-backup based algorithm is
found superior to all other methods. רוצה לראות פחות

Universal compression algorithms can detect recurring patterns in any
type of temporal data—including financial data—for the purpose of compression.
The universal algorithms actually find a model of the data that can be used for either
compression or prediction. We present a universal Variable Order Markov (VOM)
model and use it to test the weak form of the Efficient Market Hypothesis (EMH).
The EMH is tested for 12 pairs of international intra-day currency exchange rates… רוצה לראות יותר

Universal compression algorithms can detect recurring patterns in any
type of temporal data—including financial data—for the purpose of compression.
The universal algorithms actually find a model of the data that can be used for either
compression or prediction. We present a universal Variable Order Markov (VOM)
model and use it to test the weak form of the Efficient Market Hypothesis (EMH).
The EMH is tested for 12 pairs of international intra-day currency exchange rates for
one year series of 1, 5, 10, 15, 20, 25 and 30 min. Statistically significant compression
is detected in all the time-series and the high frequency series are also predictable
above random. However, the predictability of the model is not sufficient to generate
a profitable trading strategy, thus, Forex market turns out to be efficient, at least most
of the time. רוצה לראות פחות

Background: The definition of a distance measure plays a key role in the evaluation of different
clustering solutions of gene expression profiles. In this empirical study we compare different
clustering solutions when using the Mutual Information (MI) measure versus the use of the well
known Euclidean distance and Pearson correlation coefficient.
Results: Relying on several public gene expression datasets, we evaluate the homogeneity and
separation scores of different clustering… רוצה לראות יותר

Background: The definition of a distance measure plays a key role in the evaluation of different
clustering solutions of gene expression profiles. In this empirical study we compare different
clustering solutions when using the Mutual Information (MI) measure versus the use of the well
known Euclidean distance and Pearson correlation coefficient.
Results: Relying on several public gene expression datasets, we evaluate the homogeneity and
separation scores of different clustering solutions. It was found that the use of the MI measure
yields a more significant differentiation among erroneous clustering solutions. The proposed
measure was also used to analyze the performance of several known clustering algorithms. A
comparative study of these algorithms reveals that their "best solutions" are ranked almost
oppositely when using different distance measures, despite the found correspondence between
these measures when analysing the averaged scores of groups of solutions.
Conclusion: In view of the results, further attention should be paid to the selection of a proper
distance measure for analyzing the clustering of gene expression data. רוצה לראות פחות

Variable order Markov models and variable order Bayesian trees have been proposed
for the recognition of cis-regulatory elements, and it has been demonstrated that they
outperform traditional models such as position weight matrices, Markov models, and
Bayesian trees for the recognition of binding sites in prokaryotes. Here, we study to which
degree variable order models can improve the recognition of eukaryotic cis-regulatory
elements. We find that variable order models can… רוצה לראות יותר

Variable order Markov models and variable order Bayesian trees have been proposed
for the recognition of cis-regulatory elements, and it has been demonstrated that they
outperform traditional models such as position weight matrices, Markov models, and
Bayesian trees for the recognition of binding sites in prokaryotes. Here, we study to which
degree variable order models can improve the recognition of eukaryotic cis-regulatory
elements. We find that variable order models can improve the recognition of binding sites
of all of the studied transcription factors. To ease a systematic evaluation of different
model combinations based on problem-specific data sets and allow genomic scans of
cis-regulatory elements based on fixed and variable order Markov models and Bayesian
trees, we provide the VOMBAT server to the public community. רוצה לראות פחות

This work proposes an approach to remote learning of robotic cells based on internet and simulation tools. The proposed approach, which integrates remote-learning and tele-operation into a generic scheme, is designed to enable students and developers to set-up and manipulate a robotic cell remotely. Its implementation is based on a dedicated website that supports a full three-dimensional simulation and a manual online control of the cell. The website provides high flexibility for making changes… רוצה לראות יותר

This work proposes an approach to remote learning of robotic cells based on internet and simulation tools. The proposed approach, which integrates remote-learning and tele-operation into a generic scheme, is designed to enable students and developers to set-up and manipulate a robotic cell remotely. Its implementation is based on a dedicated website that supports a full three-dimensional simulation and a manual online control of the cell. The website provides high flexibility for making changes in the design of the work area, independent of the user’s location. The website can be used to further integrate full-scale remote learning labs in collaboration with other organizations רוצה לראות פחות

The classical funnel experiment was used by Deming to promote the idea of statistical
process control (SPC). The popular example illustrates that the implementation of
simple feedback rules to stationary processes violates the independence assumption
and prevents the implementation of conventional SPC. However, Deming did
not indicate how to implement SPC in the presence of such feedback rules.
This pedagogical gap is addressed here by introducing a simple feedback rule to
the… רוצה לראות יותר

The classical funnel experiment was used by Deming to promote the idea of statistical
process control (SPC). The popular example illustrates that the implementation of
simple feedback rules to stationary processes violates the independence assumption
and prevents the implementation of conventional SPC. However, Deming did
not indicate how to implement SPC in the presence of such feedback rules.
This pedagogical gap is addressed here by introducing a simple feedback rule to
the funnel example that results in a nonlinear process to which the traditional SPC
methods cannot be applied. The proposed method of Markov-based SPC, which is a
simplified version of the context-based SPC method, is shown to monitor the modified
process well רוצה לראות פחות

This paper presents a method for annotating coding and noncoding DNA regions by using variable order Markov (VOM) models. A main advantage in using VOM models is that their order may vary for different sequences, depending on the sequences’ statistics. As a result, VOM models are more flexible with respect to model parameterization and can be trained on relatively short sequences and on low-quality datasets, such as expressed sequence tags (ESTs). The paper presents a modified VOM model for… רוצה לראות יותר

This paper presents a method for annotating coding and noncoding DNA regions by using variable order Markov (VOM) models. A main advantage in using VOM models is that their order may vary for different sequences, depending on the sequences’ statistics. As a result, VOM models are more flexible with respect to model parameterization and can be trained on relatively short sequences and on low-quality datasets, such as expressed sequence tags (ESTs). The paper presents a modified VOM model for detecting and correcting insertion and deletion sequencing errors that are commonly found in ESTs. In a series of experiments the proposed method is found to be robust to random errors in these sequences רוצה לראות פחות

We suggest an experimentation strategy for the robust design of empirically fitted models. The suggested approach is used to design experiments that minimize the variance of the optimal robust solution. The new design-of-experiment optimality criterion, termed Vs-optimal, prioritizes the estimation of a model’s coefficients, such that the variance of the optimal solution is minimized by the
performed experiments. It is discussed how the proposed criterion is related to known optimality… רוצה לראות יותר

We suggest an experimentation strategy for the robust design of empirically fitted models. The suggested approach is used to design experiments that minimize the variance of the optimal robust solution. The new design-of-experiment optimality criterion, termed Vs-optimal, prioritizes the estimation of a model’s coefficients, such that the variance of the optimal solution is minimized by the
performed experiments. It is discussed how the proposed criterion is related to known optimality criteria. We present an analytical formulation of the suggested approach for linear models and a numerical procedure for higher-order or nonpolynomial models. In comparison with conventional robust-design methods, our approach provides more information on the robust solution by umerically generating its multidimensional distribution. Moreover, in a case study, the proposed approach results in a better robust solution in comparison with these standard methods. רוצה לראות פחות

We present the architecture of an elementary gene-finding algorithm that is based on a Variable Order Markov model (VOM). The VOM model is a generalization of the traditional Markov model that can cope with varying memory dependencies. The VOM model is more efficient in terms of its parameterization and therefore can be trained on relatively short sequences. Experiments with the proposed gene-finder (GF) on three prokaryotic genomes indicate its potential advantage on the detection of short… רוצה לראות יותר

We present the architecture of an elementary gene-finding algorithm that is based on a Variable Order Markov model (VOM). The VOM model is a generalization of the traditional Markov model that can cope with varying memory dependencies. The VOM model is more efficient in terms of its parameterization and therefore can be trained on relatively short sequences. Experiments with the proposed gene-finder (GF) on three prokaryotic genomes indicate its potential advantage on the detection of short genes. רוצה לראות פחות

Variable order Markov models and variable order
Bayesian trees have been proposed for the recognition
of transcription factor binding sites, and it
could be demonstrated that they outperform traditional
models, such as position weight matrices,
Markov models and Bayesian trees. We develop a
web server for the recognition of DNA binding sites
based on variable order Markov models and variable
order Bayesian trees offering the following functionality:
(i) given datasets… רוצה לראות יותר

Variable order Markov models and variable order
Bayesian trees have been proposed for the recognition
of transcription factor binding sites, and it
could be demonstrated that they outperform traditional
models, such as position weight matrices,
Markov models and Bayesian trees. We develop a
web server for the recognition of DNA binding sites
based on variable order Markov models and variable
order Bayesian trees offering the following functionality:
(i) given datasets with annotated binding
sites and genomic background sequences, variable
order Markov models and variable order Bayesian
trees can be trained; (ii) given a set of trained
models, putative DNA binding sites can be predicted
in a given set of genomic sequences and (iii)
given a dataset with annotated binding sites and
a dataset with genomic background sequences,
cross-validation experiments for different model
combinations with different parameter settings can
be performed. Several of the offered services are
computationally demanding, such as genome-wide
predictions of DNA binding sites in mammalian
genomes or sets of 104
-fold cross-validation experiments
for different model combinations based on
problem-specific data sets. In order to execute these
jobs, and in order to serve multiple users at the
same time, the web server is attached to a Linux
cluster with 150 processors. רוצה לראות פחות

In certain types of processes, verification of the quality of the output units is possible only after the entire batch has been processed. We develop a model that prescribes which units should be inspected and how the units that were not inspected should be disposed of, in order to minimize the expected sum of inspection costs and disposition error costs, for processes that are subject to random failure and recovery. The model is based on a dynamic programming algorithm that has a low… רוצה לראות יותר

In certain types of processes, verification of the quality of the output units is possible only after the entire batch has been processed. We develop a model that prescribes which units should be inspected and how the units that were not inspected should be disposed of, in order to minimize the expected sum of inspection costs and disposition error costs, for processes that are subject to random failure and recovery. The model is based on a dynamic programming algorithm that has a low computational complexity. The study also includes a sensitivity analysis under a variety of cost and probability scenarios, supplemented by an analysis of the smallest batch that requires inspection, the expected number of inspections, and the performance of an easy to implement heuristic. רוצה לראות פחות

Motivation: We propose a new class of variable-order Bayesian network (VOBN) models for the identification of transcription factor binding sites (TFBSs). The proposed models generalize the widely used position weight matrix (PWM) models, Markov models and Bayesian network models. In contrast to these models, where for each position a fixed subset of the remaining positions is used to model dependencies, in VOBN models, these subsets may vary based on the specific nucleotides observed, which are… רוצה לראות יותר

Motivation: We propose a new class of variable-order Bayesian network (VOBN) models for the identification of transcription factor binding sites (TFBSs). The proposed models generalize the widely used position weight matrix (PWM) models, Markov models and Bayesian network models. In contrast to these models, where for each position a fixed subset of the remaining positions is used to model dependencies, in VOBN models, these subsets may vary based on the specific nucleotides observed, which are called the context. This flexibility turns out to be of advantage for the classification and analysis of TFBSs, as statistical dependencies between nucleotides in different TFBS positions (not necessarily adjacent) may be taken into account efficiently—in a position-specific and context-specific manner.
Results: We apply the VOBN model to a set of 238 experimentally verified sigma-70 binding sites in Escherichia coli. We find that the VOBN model can distinguish these 238 sites from a set of 472 intergenic ‘nonpromoter’ sequences with a higher accuracy than fixed-order Markov models or Bayesian trees. We use a replicated stratified-holdout experiment having a fixed true-negative rate of 99.9%. We find that for a foreground inhomogeneous VOBN model of order 1 and a background
homogeneous variable-order Markov (VOM) model of order 5, the obtained mean true-positive (TP) rate is 47.56%. In comparison, the best TP rate for the conventional models is 44.39%, obtained from a foreground PWM model and a background 2nd-order Markov model. As the standard deviation of the estimated TP rate is 0.01%, this improvement is highly significant.
רוצה לראות פחות

—In this paper, we suggest a potential use of data compression
measures, such as the Entropy, and the Huffman Coding,
to assess the effects of noise factors on the reliability of tested systems.
In particular, we extend the Taguchi method for robust design
by computing the entropy of the percent contribution values
of the noise factors. The new measures are computed already at
the parameter-design stage, and together with the traditional S/N
ratios enable the specification… רוצה לראות יותר

—In this paper, we suggest a potential use of data compression
measures, such as the Entropy, and the Huffman Coding,
to assess the effects of noise factors on the reliability of tested systems.
In particular, we extend the Taguchi method for robust design
by computing the entropy of the percent contribution values
of the noise factors. The new measures are computed already at
the parameter-design stage, and together with the traditional S/N
ratios enable the specification of a robust design. Assuming that
(some of) the noise factors should be naturalized, the entropy of a
design reflects the potential efforts that will be required in the tolerance-design
stage to reach a more reliable system. Using a small
example, we illustrate the contribution of the new measure that
might alter the designer decision in comparison with the traditional
Taguchi method, and ultimately obtain a system with a lower
quality loss.
Assuming that the percent contribution values can reflect the
probability of a noise factor to trigger a disturbance in the system
response, a series of probabilistic algorithms can be applied to the
robust design problem. We focus on the Huffman coding algorithm,
and show how to implement this algorithm such that the designer
obtains the minimal expected number of tests in order to find the
disturbing noise factor. The entropy measure, in this case, provides
the lower bound on the algorithm’s performance. רוצה לראות פחות

Changes in stochastic processes often affect their description length, and reflected by their stochastic complexity measures. Monitoring the stochastic complexity of a sequence (or, equivalently, its code length) can detect process changes that may be undetectable by traditional SPC methods. The context tree is proposed here as a universal compression algorithm for measuring the stochastic complexity of a state-dependent discrete process. The advantage of
the proposed method is in the… רוצה לראות יותר

Changes in stochastic processes often affect their description length, and reflected by their stochastic complexity measures. Monitoring the stochastic complexity of a sequence (or, equivalently, its code length) can detect process changes that may be undetectable by traditional SPC methods. The context tree is proposed here as a universal compression algorithm for measuring the stochastic complexity of a state-dependent discrete process. The advantage of
the proposed method is in the reduced number of samples that are needed for reliable monitoring.
רוצה לראות פחות

The term throughput, which is commonly used as a performance measure of various production systems, has not been uniquely defined for multiple part-type systems. In some cases, the analytical procedures that were developed to maximize throughput of multiple part-type systems are difficult to evaluate and justify. Moreover, in some cases an inaccurate definition of this term motivated incorrect operational concepts. This paper discusses some of the problems in the traditional definition of… רוצה לראות יותר

The term throughput, which is commonly used as a performance measure of various production systems, has not been uniquely defined for multiple part-type systems. In some cases, the analytical procedures that were developed to maximize throughput of multiple part-type systems are difficult to evaluate and justify. Moreover, in some cases an inaccurate definition of this term motivated incorrect operational concepts. This paper discusses some of the problems in the traditional definition of throughput and suggests a new, more suitable definition for a multiple-product manufacturing system. רוצה לראות פחות

In this paper we follow previous “pseudo-stochastic” approaches that solve stochastic control problems by using deterministic optimal control methods. In a similar manner to the certainty equivalence principle, the suggested model maximizes a given profit function of the expected system outcome. However, unlike the certainty equivalence principle, we model the expected influences of all future events (including those that are expected beyond the planning horizon), as encapsulated by their… רוצה לראות יותר

In this paper we follow previous “pseudo-stochastic” approaches that solve stochastic control problems by using deterministic optimal control methods. In a similar manner to the certainty equivalence principle, the suggested model maximizes a given profit function of the expected system outcome. However, unlike the certainty equivalence principle, we model the expected influences of all future events (including those that are expected beyond the planning horizon), as encapsulated by their density functions and not only by their mean values. The model is applied to the optimal scheduling of multiple part-types on a single machine that is subject to random
failures and repairs. The objective of the scheduler is to maximize the profit function of the produced multiple-part mix. A numerical study is performed to evaluate the suggested pseudo-stochastic solutions under various conditions. These solutions are compared to a profit upper bound of the stochastic optimal control solutions רוצה לראות פחות

This paper presents the performance of the Weight-Balanced Testing (WBT)algorithm with multiple testers. The WBT algorithm aims to minimize the expected number of (round of)tests and has been proposed for coding, memory storage, search and testing applications. It often provides reasonable results if used with a single tester. Yet, the performance of the WBT algorithm with multiple testers and particularly its upper bound have not been previously analyzed, despite the large body of literature… רוצה לראות יותר

This paper presents the performance of the Weight-Balanced Testing (WBT)algorithm with multiple testers. The WBT algorithm aims to minimize the expected number of (round of)tests and has been proposed for coding, memory storage, search and testing applications. It often provides reasonable results if used with a single tester. Yet, the performance of the WBT algorithm with multiple testers and particularly its upper bound have not been previously analyzed, despite the large body of literature that exists on the WBT algorithm, and the recent papers that suggest its use in various testing applications. Here we demonstrate that WBT algorithm with multiple testers is far from being the optimal search procedure. The main result of this paper is the generalization of the upper bound on the expected number of tests previously obtained for a single-tester WBT algorithm. For this purpose, we first draw an analogy between the WBT algorithm and alphabetic codes; both being represented by the same Q-ary search tree. The upper bound is then obtained on the expected path length of a Q-ary tree, which is constructed by the WBT algorithm. Applications to the field of testing and some numerical examples are presented for illustrative purposes. רוצה לראות פחות

Conventional Statistical Process Control (SPC) schemes fail to monitor nonlinear and finite-state processes that often result from feedback-controlled processes. SPC methods that are designed to monitor autocorrelated processes usually assume a known model (often an ARIMA) that might poorly describe the real process. In this paper, we present a novel SPC methodology based on context
modeling of finite-state processes. The method utilizes a series of context-tree models to estimate the… רוצה לראות יותר

Conventional Statistical Process Control (SPC) schemes fail to monitor nonlinear and finite-state processes that often result from feedback-controlled processes. SPC methods that are designed to monitor autocorrelated processes usually assume a known model (often an ARIMA) that might poorly describe the real process. In this paper, we present a novel SPC methodology based on context
modeling of finite-state processes. The method utilizes a series of context-tree models to estimate the conditional distribution of the process output given the context of previous observations. The Kullback-Leibler divergence statistic is derived to indicate significant changes in the trees along the process. The method is implemented in a simulated flexible manufacturing system in order to detect significant changes in its production mix ratio output. רוצה לראות פחות

Most statistical process control (SPC) methods are not suitable for monitoring nonlinear and statedependent processes. This article introduces the context-based SPC (CSPC) methodology for statedependent data generated by a finite-memory source. The key idea of the CSPC is to monitor the statistical attributes of a process by comparing two context trees at any monitoring period of time. The first is a reference tree that represents the “in control” reference behavior of the process; the second… רוצה לראות יותר

Most statistical process control (SPC) methods are not suitable for monitoring nonlinear and statedependent processes. This article introduces the context-based SPC (CSPC) methodology for statedependent data generated by a finite-memory source. The key idea of the CSPC is to monitor the statistical attributes of a process by comparing two context trees at any monitoring period of time. The first is a reference tree that represents the “in control” reference behavior of the process; the second is a monitored tree, generated periodically from a sample of sequenced observations, that represents the behavior of the process at that period. The Kullback–Leibler (KL) statistic is used to measure the relative “distance” between these two trees, and an analytic distribution of this statistic is derived. Monitoring the KL statistic indicates whether there has been any significant change in the process that requires intervention. An example of buffer-level monitoring in a roduction system demonstrates the viability of the new method with respect to conventional methods רוצה לראות פחות

In this paper we present a novel treatment of the inspection-system design problem when inspection is unreliable and subject to classification errors. Our approach, based on the theory of Error-Correcting Codes (ECC), leads to the development of a SelfCorrecting Inspection (SCI) decision rule that does not require complete knowledge of inspection error probabilities. We show that the proposed rule assures correct classification, if the number of inspection errors is less than a certain number… רוצה לראות יותר

In this paper we present a novel treatment of the inspection-system design problem when inspection is unreliable and subject to classification errors. Our approach, based on the theory of Error-Correcting Codes (ECC), leads to the development of a SelfCorrecting Inspection (SCI) decision rule that does not require complete knowledge of inspection error probabilities. We show that the proposed rule assures correct classification, if the number of inspection errors is less than a certain number. We analyze the performance of the SCI decision rule under different inspection situations, including some situations that are uncommon in the field of error-correcting codes. Then, we show how the underlying mathematical structure can be applied to determine the number
of inspections and the level of inspection reliability in order to minimize the sum of inspection-related costs. The practical contribution of this work lies in that it expands the ability of the designer of inspection systems to deal with cases where there is very little or no information regarding the reliability of the inspection operations. רוצה לראות פחות

The paper considers a sequential Design Of Experiments (DOE) scheme. Our objective is to maximize both information and economic measures over a feasible set of experiments. Optimal DOE strategies are developed by introducing information criteria based on measures adopted from information theory. The evolution of acquired information along various stages of experimentation
is analyzed for linear models with a Gaussian noise term. We show that for particular cases, although the amount… רוצה לראות יותר

The paper considers a sequential Design Of Experiments (DOE) scheme. Our objective is to maximize both information and economic measures over a feasible set of experiments. Optimal DOE strategies are developed by introducing information criteria based on measures adopted from information theory. The evolution of acquired information along various stages of experimentation
is analyzed for linear models with a Gaussian noise term. We show that for particular cases, although the amount of
information is unbounded, the desired rate of acquiring information decreases with the number of experiments. This observation implies that at a certain point in time it is no longer efficient to continue experimenting. Accordingly, we investigate methods of stochastic dynamic programming under imperfect state information as appropriate means to obtain optimal experimentation policies. We propose cost-to-go functions that model the trade-off between the cost of additional experiments and the benefit of incremental information. We formulate a general stochastic dynamic programming framework for design of experiments and illustrate it by analytic and numerical implementation examples. רוצה לראות פחות

The objective of design of experiments (DOE) is addressed by introducing an information
optimality criterion, which is based on concepts adopted from information theory. In particular,
experiments are specied to maximize the information in the system responses about estimators of
the system parameters. It is shown that one has to maintain certain resolution of the design matrix
to maximize the information, obtainable by a design, about a system described by a linear model
with… רוצה לראות יותר

The objective of design of experiments (DOE) is addressed by introducing an information
optimality criterion, which is based on concepts adopted from information theory. In particular,
experiments are specied to maximize the information in the system responses about estimators of
the system parameters. It is shown that one has to maintain certain resolution of the design matrix
to maximize the information, obtainable by a design, about a system described by a linear model
with interactions. The correspondence between error-correcting codes and fractional factorial
experiments provides a method to attain the required resolution with a smaller fractional factorial
experiment by increasing the number of levels associated with each factor – a result that in the
context of experimental design seems counterintuitive. In particular, the Gilbert–Varshamov and
the Singleton bounds are employed to obtain bounds on the size of the fractional experiment.
Analytical approximations and numerical results are given and illustrated by examples.
רוצה לראות פחות

In this paper, an information theoretic approach is applied to analyze the performance of a decentralized control system. The control system plays the role of a correcting device which decreases the uncertainties associated with state variables of a production line by applying an appropriate ``correcting signal'' for each deviation from the target. In particular, a distributed feedback control
policy is considered to govern a transfer production line, which consists of machines and bu€ers… רוצה לראות יותר

In this paper, an information theoretic approach is applied to analyze the performance of a decentralized control system. The control system plays the role of a correcting device which decreases the uncertainties associated with state variables of a production line by applying an appropriate ``correcting signal'' for each deviation from the target. In particular, a distributed feedback control
policy is considered to govern a transfer production line, which consists of machines and bu€ers and processes a single part type in response to a stochastic demand. It is shown how the uncertainty of the demand propagates dynamically into the production system, causing uncertainties associated with bu€er levels and machine production rates. The paper proposes upper estimates for these uncertainties as functions of the demand variance, parameters of the distributed controllers and some physical properties of the production line. The bounds are based on dynamic entropy measures of the system state and the control variables. Some practical implications into the area of decentralized controller design are proposed, an information-economical analysis is presented and a numerical study is performed. רוצה לראות פחות

Many studies have explored spreading and diffusion through complex networks. The following study examines a specific case of spreading of opinions in modern society through two spreading schemes – defined as being either through ‘word-of-mouth’ (WOM), or through online search engines (WEB). We apply both modelling and real experimental results and compare the opinions people adopt through an exposure to their friend`s opinions, as opposed to the opinions they adopt when using a search engine… רוצה לראות יותר

Many studies have explored spreading and diffusion through complex networks. The following study examines a specific case of spreading of opinions in modern society through two spreading schemes – defined as being either through ‘word-of-mouth’ (WOM), or through online search engines (WEB). We apply both modelling and real experimental results and compare the opinions people adopt through an exposure to their friend`s opinions, as opposed to the opinions they adopt when using a search engine based on the PageRank algorithm. A simulated study shows that when members in a population adopt decisions through the use of the WEB scheme, the population ends up with a few dominant views, while other views are barely expressed. In contrast, when members adopt decisions based on the WOM scheme, there is a far more diverse distribution of opinions in that population. The simulative results are further supported by an online experiment which finds that people searching information through a search engine end up with far more homogenous opinions as compared to those asking their friends. רוצה לראות פחות

Irad wrote and edited six books, has published more than 80 scientific papers and patents, received numerous best papers awards and supervised more than 40 graduate students.

For Irad's selected publications list please see http://www.eng.tau.ac.il/~bengal/publications.html

This paper addresses the problem of optimizing a factorial system with multiple
responses. A heuristic termed probabilistic sequential methodology (PSM) is proposed.
The PSMidenti®es those designs that maximize the likelihood of satisfying
a given set of functional requirements. It is based on sequential experimentation,
statistical inference and a probabilistic local search. The PSM comprises three
main steps: (1) screening and estimating the main location and dispersion e€… רוצה לראות יותר

This paper addresses the problem of optimizing a factorial system with multiple
responses. A heuristic termed probabilistic sequential methodology (PSM) is proposed.
The PSMidenti®es those designs that maximize the likelihood of satisfying
a given set of functional requirements. It is based on sequential experimentation,
statistical inference and a probabilistic local search. The PSM comprises three
main steps: (1) screening and estimating the main location and dispersion e€ ects
by applying fractional factorial experiments (FFE) techniques; (2) based on these
e€ ects, establishing probabilistic measures for di€ erent combinations of factorlevels;
and (3) constructing a set of candidate designs from which the best solution
is selected by applying a heuristic local search. The PSM is attractive when the
exact analytic relationship between factor-level combinations and the system’s
responses is unknown; when the system involves qualitative factors; and when
the number of experiments is limited. The PSM is illustrated by a detailed case
study of a Flexible Manufacturing Cell (FMC) design.
רוצה לראות פחות

The method of robust design has long been used for the design of systems that are insensitive to noises. In this paper it is demonstrated how this approach can be used to obtain a robust eco-design (ecological design). In a case study, robust design principles are applied to the design of a factory smokestack, using the Gaussian Plume Model (GPM). The GPM is a well-known model for describing
pollutant dispersal from a point source, subject to various atmospheric conditions. In this research,… רוצה לראות יותר

The method of robust design has long been used for the design of systems that are insensitive to noises. In this paper it is demonstrated how this approach can be used to obtain a robust eco-design (ecological design). In a case study, robust design principles are applied to the design of a factory smokestack, using the Gaussian Plume Model (GPM). The GPM is a well-known model for describing
pollutant dispersal from a point source, subject to various atmospheric conditions. In this research, the mean-square-error (MSE) of the accumulated and the maximum pollution values around a given target are defined as the performance measures and used to adjust the design parameters. Both analytical and numerical approaches are used to evaluate the MSE measures over the design space. It is demonstrated how to use the non-linearity in the GPM to reach a low MSE value that produces a cheaper design configuration. The differences between the manufacturer viewpoint and the environmentalist viewpoint with respect to the considered eco-design problem are discussed and analyzed. רוצה לראות פחות