Module code: ECS659A
Credits: 15.0
Semester: SEM2
Contact: Dr Georgios Tzimiropoulos
The module covers the theoretical underpinnings and practical applications of Neural Networks and automatic differentiation as a tool for modern AI. Neural Networks & Deep Learning are now the method of choice for solving various Machine Learning problems. They are applied to several real-world problems not only within Academia but most importantly within Industry. Knowledge of Neural Networks and how to apply them to solve practical problems is now considered one of the most essential skills in the job market for a CS graduate. The module will include a detailed exposition for Neural Networks and their implementation using a Deep Learning framework. Topics covered include but not limited to: Automatic Differentiation, Stochastic Gradient Descent, Regression, Softmax Regression, Multi-Layer Perceptrons, Training of Neural Networks and hyper-parameter optimization, Convolutional Neural Networks, Recurrent Neural Networks. Applications of Neural Networks to Vision and NLP.
Connected course(s): UDF DATA
Assessment: 50.0% Practical, 50.0% Examination
Level: 6
