Machine Learning and Knowledge Discovery

This course is designed to provide students with a background in fundamental and advanced concepts, tools and methodology in machine learning as well as their applicability to real world problems. An overview of algorithms used in machine learning and machine learning models for supervised (classification, regression) and unsupervised learning (clustering), feature selection and dimensionality reduction, error estimation and empirical validation will be introduced. Advanced concepts such as deep feed forward neural networks and back propagation, regularization, activation functions, loss function, batch normalization as well as key deep network architectures (convolutional neural networks. auto encoders, recurrent neural network, long short-term memory (LSTM) networks) will be discussed. Students will gain hands-on experience in using various software packages and tools.

031285