APS360-Applied-Fundamentals-of-Machine-Learning
Description
Course labs of APS360 (Winter 2020) at the University of Toronto.
Lab1:
-Perform basic PyTorch tensor operations;
-Load data into PyTorch;
-Configure an Artificial Neural Network (ANN) using PyTorch;
-Train ANNs using PyTorch;
-Evaluate different ANN configuations.
Lab2:
-Understand at a high level the training loop for a machine learning model;
-Understand the distinction between training, validation, and test data;
-Understand the concepts of overfitting and underfitting;
-Investigate how different hyperparameters, such as learning rate and batch size, affect the success of training;
-Compare an ANN (aka Multi-Layer Perceptron) with a CNN.
Lab3:
-Generate and preprocess my own data;
-Load and split data for training, validation and testing;
-Train a Convolutional Neural Network;
-Apply transfer learning to improve my model.
Lab4:
-Clean and process continuous and categorical data for machine learning;
-Implement an autoencoder that takes continuous and categorical (one-hot) inputs;
-Tune the hyperparameters of an autoencoder;
-Use baseline models to help interpret model performance.
Lab5:
-Clean and process text data for machine learning;
-Understand and implement a character-level recurrent neural network;
-Use torchtext to build recurrent neural network models;
-Understand batching for a recurrent neural network, and use torchtext to implement RNN batching.
Contact
If you find any bugs in my code, or you have any questions, or you just want to chat, please do not hesitate to contact me :)
Email: nini.li@mail.utoronto.ca
or WeChat: Ljn13951644751