Intellegens’ Post

🤔 Do you know about Design of Experiments (DoE) ? ✍ In the world of experimental research, how can we optimize processes and products effectively? Enter Design of Experiments (DoE), a robust statistical methodology that transforms our understanding of the relationships between input variables (factors) and output outcomes (responses). 🔍 But what exactly does this entail? How do we evaluate the effectiveness of our experimental design model? 💡 Today’s reading includes the common terms of DoE: Design; Factors and Levels; Responses; Analysis of Variance (ANOVA); R² and Adjusted R²; Residual Analysis and Lack of Fit, etc. ✍ Curious about how these elements come together to ensure your experiments are both accurate and reliable? What insights could effective DoE bring to your work? Let’s explore together! 👉 Read the full article here: https://lnkd.in/gEyN4V7y

Fascinating article on the various techniques for studying the future! Here are some key points that caught my attention: - Backcasting: A method that focuses on envisioning preferred futures first, independent of the path to get there. - Delphi Analysis: Utilizing expert opinions versus general population views for a more focused approach. - Focus Groups: Bringing together a diverse small group for discussions on a specific theme without a structured agenda. - Technological Forecasting: Narrow expert-based forecasting on specific areas like technological advancements. - Models or Simulations: Using mathematical relationships to predict future scenarios based on system understanding. These techniques offer unique perspectives for understanding potential futures. What are your thoughts on using such methods to anticipate what lies ahead? https://lnkd.in/e89rhyvU

At JMP, we are proud to have experts like Phil Kay, who consistently share their knowledge on Design of Experiments (DOE). In his LinkedIn post, Phil breaks down the concept of correlation maps offering valuable insights for professionals and enthusiasts alike. Don't forget to follow his series #DOEbyPhilKay for more expert advice and tips. Read on to expand your understanding of correlation maps and how they can improve your experimental designs

📃Scientific paper: Statistically Distinct Plans for Multi-Objective Task Assignment Abstract: We study the problem of finding statistically distinct plans for stochastic planning and task assignment problems such as online multi-robot pickup and delivery (MRPD) when facing multiple competing objectives. In many real-world settings robot fleets do not only need to fulfil delivery requests, but also have to consider auxiliary objectives such as energy efficiency or avoiding human-centered work spaces. We pose MRPD as a multi-objective optimization problem where the goal is to find MRPD policies that yield different trade-offs between given objectives. There are two main challenges: 1) MRPD is computationally hard, which limits the number of trade-offs that can reasonably be computed, and 2) due to the random task arrivals, one needs to consider statistical variance of the objective values in addition to the average. We present an adaptive sampling algorithm that finds a set of policies which i) are approximately optimal, ii) approximate the set of all optimal solutions, and iii) are statistically distinguishable. We prove completeness and adapt a state-of-the-art MRPD solver to the multi-objective setting for three example objectives. In a series of simulation experiments we demonstrate the advantages of the proposed method compared to baseline approaches and show its robustness in a sensitivity analysis. The approach is general and could be adapted to other multi-objective task assignment and planning problems under uncertainty. Continued on ES/IODE ➡️ https://etcse.fr/N4aU ------- If you find this interesting, feel free to follow, comment and share. We need your help to enhance our visibility, so that our platform continues to serve you.

"Exciting research update! 🚀 Our paper on "Analyzing the fuzzy reliability of a skim milk powder production system using intuitionistic fuzzy number" has been published in Mathematics in Engineering, Science and Aerospace (August 2022). 📄 In this study, we employed the universal generating function (UGF) to assess the reliability of a complex manufacturing system with Computer Numerical Control (CNC). We also evaluated various configurations of sub-systems, including series, parallel, and mixed arrangements. This research was made possible under the guidance of Dr. Akshay Swami. 🙏 Want to learn more? Access the paper here: [ https://rb.gy/o95eqe ] #researchupdate #reliabilityengineering #fuzzymathematics #manufacturingsystems #publication"

I am happy to share that my first publication appeared in the special issue of the Journal of Economic Behavior & Organization. In a joint paper with Cars Hommes, Stefanie J. Huber, and Isabelle Salle, we study group coordination and price dynamics in complex environments using a lab experiment. This paper highlights three key findings. First, the price converges to the simplest equilibria in all markets. Second, we document a novel and intriguing finding: there is a non-monotonicity of the behavior when complexity increases. Convergence to the two-cycle occurs for the intermediate parameter range, while both high- and low-complexity scenarios lead to coordination on the simplest equilibrium (steady state) in the lab. All indicators of coordination and convergence significantly exhibit this non-monotonic relationship in the learning-to-forecast experiments and this non-monotonicity persists in the learning-to-optimize design. Third, convergence in the learning-to-optimize experiment is more challenging to achieve: coordination on the two-cycle is never observed, although the two-cycle Pareto dominates the steady state in our design. Read the full paper at https://lnkd.in/dfY9NCkj.

What to know the real impact of Digital Twins? If you are working on digital twins and looking to understand applications and its different aspects, check this paper by Francisco Gómez Medina, PhD and Veronica Martínez - Well done both!!!! #datastrategy #digitaltwins #leadership #emergingtechnologies

Representation Alignment for Generation: Training Diffusion Transformers Is Easier Than You Think #DiffusionTransformers #AIResearch #DeepLearning #GenerativeModels #TransformerArchitecture #MachineLearning #ComputerVision #ArtificialIntelligence #NaturalLanguageProcessing #AIInnovation https://sihyun.me/REPA/

🔍 Why use more factors at the beginning of DoE? 💡 Design of Experiments (DoE) expert Bradley Jones sheds light on reasons why DoE may appear to fall short of expectations. Why do some people mistakenly think that DoE is either not suitable for their needs or too expensive to implement? One reason that experiments might fail is the opportunity space for success in DoE, which increases with more factors, especially at the beginning. Design of Experiments (DoE) focuses on optimization, aiming for excellent product and process characteristics. Do not settle for mediocracy. Finally, bad experiences with DoE often occur, particularly with continuous factors, when these factors aren't varied enough. Using small ranges can lead to missing the optimal conditions of the process. 💡 In our 4-part series we talked with Bradley Jones and shared the most valuable insights on the importance and positive impact of DoE on research, development and production processes. 🎯 Head over to the full article and discover how you can improve with DoE: https://lnkd.in/eWWAGdX2 #DoE #optimalprocesses #BradleyJones #Effex

Excited to share our latest research on operator learning using function encoders! In our new paper, "Basis-to-Basis Operator Learning Using Function Encoders," we introduce a novel approach for learning operators on Hilbert spaces of functions. Our Basis-to-Basis (B2B) approach works in two main stages: 1. We learn sets of basis functions to efficiently represent both input and output function spaces, using function encoders (neural network-parameterized basis functions). 2. We then learn an operator that maps between these function representations. We propose three variants: Basis-to-Basis (B2B), Singular Value Decomposition (SVD), and Eigendecomposition (ED) Some key highlights: - We provide theoretical guarantees on generalization for linear operators - We provide detailed comparisons against state-of-the-art methods like DeepONet on several benchmark problems, including Darcy flow and linear elasticity - Our approach can handle variable input sampling locations, unlike many existing methods This work opens up new possibilities for efficiently modeling complex physical systems and PDEs. I'm particularly excited about the potential fluid dynamics, materials science, and climate modeling applications. Check out the preprint https://lnkd.in/eWHyzEZ4 and code https://lnkd.in/eSWbt4zS for all the technical details! Feedback and collaboration opportunities are welcome. #HibertSpaceLearning #ScientificMachineLearning #OperatorLearning #Basis2BasisOperatorLearning #NeuralOperators #AI4Science