Neural Networks and Deep Learning Practical

Department of Mathematics and Computing
IIT(ISM) Dhanbad
Course Name: Neural Networks and Deep Learning Practical

Course Code: MCC542
Lab Location: Computer Lab, M&C
Do’s and Don’ts:
1. Come to the lab. in time. After 5 minutes of entry time nobody will be allowed.
2. Follow the Instruction provided by the Faculty/Research Scholar at the beginning.
2. Write the program using the instructions given and then start coding.
3. Don’t take out the computer accessories like: mouse, patch chord, etc.
4. During the Lab. hour don’t create any disturbance
Software Requirement:
1. PYTHON
2. MATLAB
3. R
4. SPSS
1. S. N.Sivanandam and S. N.Deepa, Introduction to neural networks using Matlab, 2016.

2. Simon Haykin, Neural Networks and Learning Machines, PHI, 2008.
Reference books:
Lab Examination:
After completion of all the programs and project work in the given list of experiments there
will be one lab exam of 100 marks during the slot provided in the academic calendar.
Examination will be based on Programming followed by Viva Voce.
The 100 marks will distributed as follows:
60 Marks in program writing and execution, 30 marks in Lab. File, 10 marks in Viva Voce.
LIST OF PROGRAMS
1. Implementation of different activation functions to train Neural Network.

2. Implementation of different Learning Rules.
3. Implementation of Perceptron Networks.
4. Implementation of Adeline network for system identification.
5. Implementation of Madeline network
6. Pattern matching using different rules.
7. Project related to application of machine learning in healthcare.
8. Project related to application of machine learning in business analysis.
9. Project related to application of machine learning in sports analytics
10. Project related to application of machine learning in Time Series Analysis &
Forecasting.
EXPERIMENT-1
Aim: Implementation of different activation functions to train Neural Network.

Tools/Apparatus: None
Procedure:
1) Check data that is linearly separable or not.
2) Analyze the activation functions.
3) Set up code for plotting.
EXPERIMENT-2
Aim: Implementation of different Learning Rules.

Procedure:
1) Initialize the input vectors associated with the target values.
2) Initialize the weights and bias.
3) Set learning rule.
4) Input layer has unique activation function.
5) Calculate the output.
6) Make adjustment of weights comparing the desired output and target values.
7) Continue the iterations until there is no change of weights.
EXPERIMENT-3
Aim: Implementation of Perceptron Networks

Procedure:
1) Initialize the input vector.
2) Train the network weights for the perceptron.
3) Make predictions with the perceptron.
EXPERIMENT-4
Aim: Implementation of Adeline network for system identification

Procedure:
1) During the training of the Adaline network give the input and desired output(D) as it is a
supervised learning.
2) Calculate network output(Y) by the formula as follows
Y = ΣXi(Wi)
T = X1W1+ X2W2+X3W3+…..XnWn
3) Compute error as
Error (e) = (D-Y)
4) Adjust the weights by delta rule as follows
Wnew = Wold + α (D-Y)Xi
(Wnew - Wold ) = α (D-Y)Xi
∆W = α (D-Y)Xi, Where
Where α= learning rate; D= desired output; Y= network output; Xi= input, ∆W= change
in weight.
EXPERIMENT-5
Aim: Implementation of Madeline network

Procedure:
1) Initialize the weights, Bias, Learning rate(α) to start the training

2) Continue step 3-8 when the stopping condition is not true.
3) Continue step 4-6 for every bipolar training pair.
4) Activate each input unit
5) Obtain the net input at each hidden layer, i.e. the Adaline layer
6) Apply the activation function to obtain the final output at the Adaline and the Madaline
layer .
7 ) Calculate the error and adjust the weights.
8 ) Test for the stopping condition, which will happen when there is no change in weight
or the highest weight change occurred during training is smaller than the specified
tolerance.
EXPERIMENT-6
Aim: Pattern matching using different rules.

Procedure:
Depending on the label of input, type of learning( i.e . supervised/ unsupervised) use any
Pattern matching algorithm where different rules can be used.
EXPERIMENT-7 / Project 1
Aim: Project related to application of machine learning in healthcare.

Procedure:
1) Identifying and diagnosing diseases and other medical issues.
2) Collection of data sets and analyzes it.

3) Use machine learning techniques for prediction and require development.
Examples:
1) Enhance medical diagnostics.

2) Drug discovery and development.
3) Automation
4) Radiation treatment.
Aim: Project related to application of machine learning in business analysis.

Procedure:
1) Identifying and understanding the market.
2) Collection of previous existing data sets and analyzes it.
3) Use machine learning techniques for prediction and for development.
Examples:
1) Stock market.
2) Market prediction.
Aim: Project related to application of machine learning in sports analytics

Procedure:
1) Initially, identify the required demand in the market and software specifications of the
system.
2) Use previously collected data for better understanding.
3) Use machine learning techniques on the data set to find meaningful patterns from
which future events can be predicted or classified.
Example:
1) Analyzing and predicting the performance of a team or individual players and
strategy.
2) Optimization of player positioning for team sports.
Aim: Project related to application of machine learning in Time Series Analysis &
Forecasting
Procedure:
1) Initially, identify the required demand in the market and software specifications of the
system.
2) Use previously collected data for better understanding.
3) Use machine learning techniques on the data set to find meaningful patterns from
which future events can be predicted or classified.
Example:
1) Business Forecasting
2) Sensor Data Monitoring
EXPERIMENT-10a / Project 6
Aim: Design of Test Suits for White Box Testing

Procedure:
1) Identify the feature, component, program to be tested.
2) Plot all possible paths in a flow graph.
3) Identify all possible paths from the flow graph.
4) Write Test Cases to cover every single path on the flow graph.
5) Execute, rinse, repeat.
Example:
White Box penetration techniques: In this testing, the tester/developer has full information
of the application's source code, detailed network information, IP addresses involved and all
server information the application runs on. The aim is to attack the code from several angles
to expose security threats
EXPERIMENT-1b / Project 7
Aim: Design of Test Suits for White Box Testing

Procedure:
1) Identify the feature, component, program to be tested.
2) Plot all possible paths in a flow graph.
3) Identify all possible paths from the flow graph.
4) Write Test Cases to cover every single path on the flow graph.
5) Execute, rinse, repeat.
Example:
White Box Mutation Testing: Mutation testing is often used to discover the best coding
techniques to use for expanding a software solution.

Neural Networks and Deep Learning Practical

Uploaded by

Copyright:

Available Formats

Neural Networks and Deep Learning Practical

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Neural Networks and Deep Learning Practical

Uploaded by

Copyright:

Available Formats

Department of Mathematics and Computing

Course Name: Neural Networks and Deep Learning Practical

2. Follow the Instruction provided by the Faculty/Research Scholar at the beginning.

4. During the Lab. hour don’t create any disturbance

1. S. N.Sivanandam and S. N.Deepa, Introduction to neural networks using Matlab, 2016.

1. Implementation of different activation functions to train Neural Network.

3. Implementation of Perceptron Networks.

4. Implementation of Adeline network for system identification.

5. Implementation of Madeline network

6. Pattern matching using different rules.

7. Project related to application of machine learning in healthcare.

8. Project related to application of machine learning in business analysis.

9. Project related to application of machine learning in sports analytics

Aim: Implementation of different activation functions to train Neural Network.

Aim: Implementation of different Learning Rules.

Aim: Implementation of Perceptron Networks

Aim: Implementation of Adeline network for system identification

Aim: Implementation of Madeline network

1) Initialize the weights, Bias, Learning rate(α) to start the training

Aim: Pattern matching using different rules.

Aim: Project related to application of machine learning in healthcare.

2) Collection of data sets and analyzes it.

1) Enhance medical diagnostics.

Aim: Project related to application of machine learning in business analysis.

1) Identifying and understanding the market.

2) Collection of previous existing data sets and analyzes it.

3) Use machine learning techniques for prediction and for development.

Aim: Project related to application of machine learning in sports analytics

Aim: Design of Test Suits for White Box Testing

Aim: Design of Test Suits for White Box Testing

You might also like