HW1

My homework report is named as HW1_Report_fergujp.pdf and is in this folder. My code is called HW1.py and is in this folder.

TO RUN THE CODE: You need torch, matplotlib, scikit-learn, pandas, and numpy(although this is installed as a dependency for other libaries) installed and python3 installed. At the bottom you can add your tests under the if name=="__main": statement, or if you import just use test_model(params)

The params are as follows:

  model_architexture = 
    Choose between "Linear Regression", "DNN-16", "DNN-30-8", "DNN-30-16-8", "DNN-30-16-8-4" and "DNN-8-4".

  activationFunction = 
    Choose between ReLU, LeakyReLU, Sigmoid, or Tanh. Do not use strings these are variables in the code

  Batchsize = 
    Choose a batch size as an int

  Learning rate =
    Choose a learning rate as an int

  Epochs = 
    Choose the number of epochs as an int

For highest-performed DNN model and Linear regression model, it will be in the code, and there will also be screenshots in /HighPerformScreenShots For screenshot's of iteration of model's training and testing with timestamps, it will be in /TrainTestScreenshots

HW2