Schedule

Day 1 - Wednesday 12th June
Model-driven AI: Logic Progamming and SAT-solving

Welcome [9:00 AM - 9:30 AM]

9:00 AM - 9:30 AM: Welcome and Overview

• Brief introduction to the day's schedule.

• Overview of AI and ML, their significance, and real-world applications.

Session 1:  Logic Programming (Part 1) [9:30 AM - 12:30 PM]

9:30 AM - 10:30 AM: Introduction to Logic Programming

• Introduction/recap of propositional and predicate logic.

• Horn clauses and clausal resolution

10:30 AM - 10:45 AM: Break

Session 2:  Logic Programming (Part 2) [9:30 AM - 12:30 PM]

10:45 AM - 12:30 PM: Logic Programming in Prolog [Practical]

• Introduction to Prolog for Logic Programming

Lunch Break [12:30 PM - 1:30 PM]

Session 3:  SAT-solving (Part 1) [1:30 PM - 4:45 PM]

1:30 PM - 3:00 PM: Introduction to SAT-solving

• The complexity of SAT

• Introduction to the DPLL algorithm for SAT-solving

3:00 PM - 3:15 PM: Break

Session 4:  SAT-solving (Part 2) [1:30 PM - 4:45 PM]

3:15 PM - 4:45 PM: Translating problems into SAT [Practical]

• Introduction to Python-sat library

• Procedurally generating and solving SAT problems

4:45 PM - 5:00 PM: Wrap-up and Conclusion

Day 2 - Thursday 13th June
Nature-Inspired AI: Randomness and Meta-Heuristics

Session 1: Introduction to Randomized Algorithms [9:00 AM - 10:30 AM]

• • Definition and importance of randomized algorithms. 

  • Randomized algorithms vs. deterministic algorithms.

  • Examples of how to design and analyse randomized methods.

Break: 10:30 AM - 10:45 AM

Session 2: Nature inspired Meta-Heuristics [10:45 AM - 12:30 PM]

• • Introduction to Local Search for combinatorial optimisation.

  • Sequential vs distributed methods.

  • Examples of the design of search methods and performance criteria.

Lunch Break: 12:30 PM - 1:30 PM

Session 3: Modelling Stochastic Processes with Markov Chains [1:30 PM - 3:00 PM]

• • Definition and basic properties of Markov chains. 

  • Transition probabilities and transition matrices.

  • Analysing Simulated Annealing with varying cooling schedules.

Break: 3:00 PM - 3:15 PM

Session 4: Real-world Applications [3:15 PM - 4:45 PM]

• • Markov chain Monte Carlo (MCMC) methods. 

  • Applications in machine learning, finance, and biology.

  • Practical examples and implementations.

Conclusion [4:45 PM - 5:00 PM]

Day 3 - Friday 14th June 2024

Data-driven AI: Clustering and Neural Networks

Morning Session: Introduction and Clustering 

9:00 AM - 10:30 AM: Foundations of Machine Learning

• Introduction to Supervised Learning: Labelled data, regression, classification, examples.

• Introduction to Unsupervised Learning: Data without labels, finding patterns, examples.

• Introduction to Reinforcement Learning: Learning through interaction with an environment, reward systems, examples.

10:30 AM - 10:45 AM: Coffee Break

Session 2: Clustering in Unsupervised Learning [10:45 AM - 12:30 PM] 

• Basics of Clustering: Definition, importance, and applications in various fields.

• Popular Clustering Algorithms: Overview of K-means, hierarchical clustering, and DBSCAN.

• Hands-On Exercise: Simple clustering exercise using a tool like Python with libraries such as scikit-learn. Participants will apply a clustering algorithm to a dataset and interpret the results.

12:30 PM - 1:30 PM: Lunch Break

Session 3: Neural Networks [1:30 PM - 4:45 PM]

1:30 PM - 3:00 PM: Introduction to Neural Networks

• Basics of Neural Networks: Neurons, weights, biases, and activation functions.

• Architecture of Neural Networks: Layers, input, hidden layers, output, how they mimic the brain.

• How Neural Networks Learn: Forward propagation, loss functions, backpropagation, and gradient descent.

3:00 PM - 3:15 PM: Break

Session 4: Deep Dive into Deep Learning [3:15 PM - 4:45 PM]

• From Neural Networks to Deep Learning: What makes a neural network 'deep', and advantages of depth.

• Convolutional Neural Networks (CNNs): Specialized for processing structured grid data such as images.

• Recurrent Neural Networks (RNNs) and LSTMs: Designed for sequential data such as time series or natural language.

• Building a Simple Neural Network: Participants will use an interactive tool or programming environment to build and train a basic neural network on a simple dataset (e.g., digit recognition using the MNIST dataset).

• Discussion and Q&A: Review the exercise results, discuss the challenges encountered, and how neural networks can be applied in various domains.

Wrap-Up and Takeaways [4:45 PM - 5:00 PM]

• Recap of the day's key points.

• Open floor for final questions and discussions.

• Guidance on further resources for deepening knowledge in ML and AI.

This schedule is designed to offer a balance between theoretical understanding and practical application, providing a solid foundation in key areas of AI and ML. Adjustments may be necessary based on the audience's background and interests, as well as logistical considerations.