Lecture
This lecture covers the basics of Multi-Layer Perceptrons (MLPs), starting with a review of logistic regression and non-separable distributions. It then delves into reformulating logistic regression, repeating the process, and introducing MLPs with hidden, input, and output layers. The lecture also discusses the binary and multi-class cases, AdaBoost, gradient descent, stochastic gradient descent, and adaptive moment estimation. It concludes with practical applications such as Rosenbrock optimization using Adaboost and MLPs, as well as the use of MLPs in solving the checkerboard problem and classifying MNIST digits.