fedme/ml-assignment-2

ML Assignment 2

Prerequisites

Python 3.7 is required together with the latest version of the following packages (all available from PIP):

• numpy
• pandas
• matplotlib
• scikitlearn
• mlrose_hiive

Obtaining the source code

1. Clone the code repository on your computer:

git clone https://github.com/fedme/ml-assignment-2

Optimization problems

1. Navigate to the directory of the optimization problem:

cd ./{continuous_peaks|four_peaks|knapsack}

2. Run the optimization for all four algorithms on the problem:

python optimization.py

This will output several stats in form of CSV files inside the stats sub-folder.

3. Run the analysis for all four algorithms on the problem:

python analysis.py

This will output several plots in form of PNG files inside the plots sub-folder.

Neural Network

1. Navigate to the directory of neural network optimization problem:

cd ./mlp

2. Run the optimization of the neural network's weights using all different algorithms:

python mlp_training.py

This will output several stats in form of CSV files inside the stats sub-folder.

3. Run the analysis of the results:

python analysis.py

This will output several plots in form of PNG files inside the plots sub-folder.