CS6482 Neural Networks

2005-2006

Pre-course work

• Pre-course work description

About the course

• About the course

Day 1: Supervised learning in neural networks

• Introduction
• Linear matrix memories
• Perceptrons and MLPs
• Generalization I

Supervised learning lab

Day 2: Applications, validation, and boosting

• Generalization II
• Applications of MLPs
• Tricks of the trade
• Committee machines and boosting

Committee machines and boosting lab

Day 3: Reinforcement learning

• Learning from immediate rewards
• Learning from delayed rewards
• Applications of reinforcement learning

Reinforcement learning lab

Day 4: Unsupervised learning

• Introduction and unsupervised clustering
• Self-organizing feature maps
• Other algorithms
• Independent component analysis

Unsupervised learning lab

Day 5: Dynamics and other topics

• The Hopfield model of memory
• Optimization by neural hardware
• A model of acquired dyslexia

Answers