- To master supervised learning techniques using deep neural networks.
- To develop advanced models for complex data sets in supervised learning.
- To explore unsupervised learning methods and their applications.
- To implement clustering and dimensionality reduction in unsupervised learning.
- To analyse data patterns using advanced unsupervised algorithms.
- To optimise neural network architectures for specific tasks.
- To evaluate deep learning models for real-world problem-solving.
UNIT 1. SUPERVISED DEEP LEARNING (I) Introduction Review: Artificial neural network (ANN) Review: ANN exercises Convolutional Neural Networks (CNN) CNN Exercises UNIT 2. SUPERVISED DEEP LEARNING (II) Natural language processing (I) Recurrent neural networks (RNN) (I) Recurrent neural networks (RNN) (II) Natural language processing (II) RNN Exercise UNIT 3. UNSUPERVISED DEEP LEARNING (I) Boltzmann Machines (BM) Restricted Boltzmann Machines (RBM) Recommender systems Recommender systems: metrics RBM exercise UNIT 4. UNSUPERVISED DEEP LEARNING (II) Self-organizing maps (SOM) SOM exercises Autoencoders (AE) AE exercises Proposed exercise