Machine Learning Applications course, spring 2024
lecture notes:
- machine learning fundamentals
- deep neural networks
- decision trees and ensemble methods
- convolutional neural networks
- computer vision: traditional methods
- computer vision: deep-learning methods
- recurrent neural networks
- attention and transformers
- autoencoders and GANs
- graph- and group-equivariant- neural networks
- self- and semi-supervised learning, synthetic data
laboratory classes:
- Preliminary problems
- simple perceptron networks
- Universal Approximation Theorem
- Colab notebook
- Handwritten digits classification using MNIST dataset with Pytorch
- models: perceptron, deep fully-connected network, generic CNN
- various activations
- overfitting
- regularization, early stopping
- Colab notebook
- ECG signal classification
- classifiers comparison: SVM, decision trees, random forests
- feature vectors
- scikit-learn library
- Colab notebook
- Deep CNNs
- VGG
- degradation problem and ResNets
- model saving and visualization
- Colab notebook
- Regularization
- L2 and L1 regularization implemented by hand
- Colab notebook
- Transformer network
- self-attention mechanism
- and Transformer encoder implemented from scratch
- Colab notebook
- Convolutional GAN on MNIST
- generative adversarial network model: generator & discriminator
- training GANs
- Colab notebook
