Glagan/42-multilayer-perceptron

The Multilayer Perceptron project at 42

Multilayer Perceptron

Requirements

Python 3 and other requirements in requirements.txt.

Usage

You first need to train a model to predict with it.
If no datasets as the first argument is given, the default one will be used.

python3 train.py

You can then predict for another dataset with the same format.
if no datasets as the first argument is given, it will try to use datasets/correction.csv.

python3 predict.py

You can see the effect of changing the learning rate and the neural network size with the learning_rate.py and network_size.py scripts, which will train a model with different hyperparameters and compare their accuracy and loss over each epoch.

python3 learning_rate.py
python3 network_size.py

Resources

• No 42-AI for this subject !
• Andrew Ng course
  • Theory: https://www.coursera.org/learn/machine-learning/home/week/4
  • Application: https://www.coursera.org/learn/machine-learning/home/week/5
• https://www.youtube.com/playlist?list=PLZHQObOWTQDNU6R1_67000Dx_ZCJB-3pi
• https://cs231n.github.io/