These notebooks are supposed to be used as presentations together with RISE extension (https://github.com/damianavila/RISE) to Jupyter.
Although the training is still under development, the expected content would be:
- Introduction (appendix for environment installation)
- Datasets - Definition, Data Visualization, Preprocessing
- Supervised Learning (featuring Linear Regression and Logistic Regression), Model evaluation
- Supervised Learning (SVMs), Kernel trick
- Supervised Learning (KNN), Hyperparameter Optimization
- Supervised Learning (Decision Trees), Handling Underfitting/Overfitting
- Supervised Learning (Random Forests), Ensembles (Bagging, Boosting, Stacking)
- Supervised Learning (Neural Network), Introduction
- Supervised Learning (Neural Network), Learning (SGD, momentum, learning rate, etc), Loss functions, Initialization, Activations (Sigmoid, tanh, ReLU, etc), Regularization (including Dropout), Batch Normalization
- Supervised Learning (Convolutional Neural Network), Introduction, R-CNN
- Supervised Learning (Deep Neural Network), Skip connections (Residual Networks), Transfer learning
- Unsupervised Learning (Kmeans), Clustering, Anomaly detection
- Unsupervised Learning (Autoencoders), Representation learning, Generative networks
- RNN, LTSM, Embedding???
- Reinforcement Learning???
- Bayesian Deep Learning???