Mathematical Sciences for Artificial Intelligence

1. Introduction to Machine Learning Applications: Overview of various machine learning applications, focusing on the issues the course aims to address: computer vision, natural language processing, recommender systems, and retrieval augmented generation. 2. Computer Vision: Study of advanced techniques for image recognition, segmentation, and video analysis. Implementation of case studies using convolutional neural networks and other deep learning techniques. 3. Natural Language Processing (NLP): Exploration of algorithms for language analysis, machine translation, speech synthesis, and text understanding. Use of models like BERT and GPT for practical applications. 4. Recommender Systems: Examination of strategies for developing personalized and scalable recommendation systems, analyzing user behavior and collaborative filtering. 5. Retrieval Augmented Generation: Introduction to models that integrate information retrieval and text generation, such as RAG and RETRO, exploring their applications in specific domains like virtual assistants and automated question answering.

1. Foundations of Machine Learning: Knowledge of basic machine learning concepts as learned in introductory courses, including supervised and unsupervised learning algorithms. 2. Programming: Strong programming skills, preferably in Python, as it is the language commonly used for machine learning libraries. 3. Mathematics and Statistics: Good command of linear algebra, differential and integral calculus, and statistics, essential for understanding and implementing machine learning algorithms. 4. Data Handling: Ability to manipulate and prepare data, understanding of issues related to data quality and its preparation for model training.

Applied Machine Learning and AI for Engineers by Jeff Prosise Released November 2022 Publisher(s): O'Reilly Media, Inc. ISBN: 9781492098058

Attendance is not mandatory

1. Project: Students will be required to develop a final project that applies machine learning techniques to one of the domains covered in the course (computer vision, natural language processing, recommender systems, or retrieval augmented generation). The project can be done individually or in small groups and must be presented at the end of the course, complete with documentation and source code. The project will be evaluated based on creativity, technical correctness, effectiveness of implementation, and quality of presentation. 2. Oral Exam: Following the project presentation, students will undergo an oral exam based on both their project and the theoretical content of the course. The oral exam will assess understanding of the algorithms, machine learning techniques used, and their ethical and social implications. It will also provide an opportunity to discuss design choices and address any doubts related to the presented project.

Lectures and exercises in the classroom