Introduction to Machine Learning

Mathematical Finance Section, Department of Mathematics, Imperial College London
Course code: M5MF45
Instructor: Justin Sirignano (j.sirignano AT

Course Topics
  • Bias-variance tradeoff, linear regression, ridge regression, lasso regression, logistic regression
  • Regularization, stochastic gradient descent, minibatch gradient descent
  • Kernel smoothing methods, Gaussian process regression
  • Decision trees, random forests, bagging, boosting, Adaboost, gradient boosting
  • Unsupervised learning, clustering, Gaussian mixtures
  • Basic neural network architectures, backpropagation
  • Universal approximation theory for neural networks
  • Deep learning, convolution neural networks, convolution filters, pooling, dropout, autoencoders, data augmentation, stochastic gradient descent with momentum (time allowing)
  • Implementation of neural networks for image classification, including MNIST and CIFAR10 datasets (time allowing)
  • Multi-armed bandits, reinforcement learning, neural networks for Q-learning (time allowing)
Useful Links Homeworks