Learning and Deep Learning

Learn to formalize a learning problem
Know the concepts of supervised and unsupervised learning, regression and classification
Understand the main learning methods (KMeans, NL-Bayes, GMM, Support Vector Machines, Deep Learning)
Know their applications
Know how to apply them.

Basic knowledge of linear algebra, probability and programming.

Introduction: Why learn? Applications


Unsupervised learning, Kmeans. Application to data clustering


Supervised learning, KNN, NL-Bayes


Support Vector Machines: linear SVMs, the Kernel Trick. Application to character recognition


Gaussian mixture model, Maximum a posteriori. Application to speaker recognition


Deep Learning: Architecture, Optimization. Stochastic Gradient Descent, DropOut, Data augmentation


Deep Learning: Convolutional neural networks. Application to character recognition


Deep Learning: Generative models, Auto-encoders, GANs. Application to denoising

Learning and Deep Learning