This three-day intensive program is designed for civil engineering and architecture students initiating their research. The course aims to provide a solid foundation in machine learning concepts and practical Python programming skills necessary for tackling complex, data-driven challenges in these domains. Participants will learn to: • Understand the core principles of machine learning and how they apply to civil engineering and architecture. • Implement machine learning algorithms in Python using popular libraries such as SciPy, Keras, and PyTorch. • Analyze and interpret results derived from real-world datasets pertinent to Civil Engineering and architecture. The course combines theoretical lectures with hands-on, guided exercises to ensure students not only grasp key concepts but also gain practical experience they can immediately apply to their research. The curriculum is designed to help participants confidently integrate machine learning methods into their doctoral work, driving innovation and solutions across a wide range of challenges in the built environment. TARGET AUDIENCE & LANGUAGE: Master students, PhD candidates, post-doctoral fellows, Researchers, and Practitioners. The course will be offered in English. COURSE MATERIALS: Lecture slides and notes. Jupiter notebooks with code examples. Access to sample datasets related to civil engineering and architecture. EXPECTED GOALS: By the end of this course, participants will be equipped with the knowledge and hands-on experience to confidently implement machine learning techniques in their research and professional projects. This course and the invited lecturers offer a rare opportunity to explore advanced research directions and emerging trends shaping the future of structural design, urban planning, construction optimization, and sustainability. Attendees will gain insight into how cutting-edge computational methods and innovations can transform traditional practices, ultimately fostering a deeper understanding of the synergy between data-driven approaches and the built environment.

Basic understanding of mathematics and statistics. Familiarity with programming concepts. Prior knowledge of Python, including packages such as Pandas, NumPy, Matplotlib, and Seaborn, is recommended. Alternatively, participants can complete a preliminary Python workshop or review the provided materials before the course begins.

Il corso di dottorato θ associato alla Summer School ARTISTE 2025 Il https://artisteconference.polito.it Day 1 – Data Exploration & General Concepts (9.5 Hours) 08:00 – 08:15 | Registration Attendees sign in and collect materials. 08:15 – 09:45 | Welcome and Course presentation: Framework, Aims, and Scope (1.5 hours) Opening remarks by course organizers. Overview of the course objectives and structure. 09:45 – 11:45 | Session 1: Machine Learning Base + Classifiers (Part I) Introduction to fundamental ML concepts (2 hours). Overview of basic classifiers and their applications. 11:45 – 12:00 | Coffee Break (15 minutes) 12:00 – 13:00 | Session 1: Classifiers (Part II) (1 hour). Continuation of classifiers. 13:00 – 14:00 | Lunch (1 hour) 14:00 – 15:30 | Session 2: Gaussian Process (Surrogate Modeling) (1.5 hours) Concepts and practical uses of Gaussian Process models in civil/architectural research. 15:30 – 15:45 | Coffee Break (15 minutes) 15:45 – 17:15 | Session 3: Data Exploration (1.5 hours) Techniques for examining and visualizing data relevant to civil engineering and architecture. 17:15 – 19:45 | Live Workshop (2 hours): Data Exploration Hands-on session applying data exploration methods to sample datasets. Guidance on best practices and troubleshooting. Day 2 – Neural Networks (10 Hours) 08:30 – 09:00 | Recap & Q&A (0.5 hours) Brief review of Day 1 content. 09:00 – 11:00 | Session 1: Introduction to Neural Networks (Part I) Fundamental concepts: perceptrons, feedforward networks, and activation functions. (2 hours) 11:00 – 11:15 | Coffee Break (15 minutes) 11:15 – 13:15 | Session 1: Introduction to Neural Networks (Part II) Training procedures, loss functions, and practical tips (2 hours). 13:15 – 14:15 | Lunch (1 hour) 14:15 – 16:15 | Session 2: Advanced Neural Networks (Part I) (2 hours). Deep Learning architectures (CNNs, RNNs), regularization, and optimization strategies. 16:15 – 16:30 | Coffee Break (15 minutes) 16:30 – 18:30 | Session 2: Advanced Neural Networks (Part II) (2 hours) Continued exploration of advanced topics (total 4 hours for the advanced segment). 18:30 – 19:30 | Live Workshop (1.5 hours): Application of Neural Networks Practical implementation of neural network models on relevant datasets. Real-time guidance for coding, training, and performance evaluation. Day 3 – Other Regression Techniques (5.5 Hours) 08:30 – 09:00 | Recap & Q&A (0.5 hours) Brief review of Day 2 content. 09:00 – 11:00 | Session 1: Tree-Based Methods & Boosting (Part I) (2 hours). Introduction to decision trees, random forests, and boosting algorithms. 11:00 – 11:15 | Coffee Break (15 minutes) 11:15 – 13:15 | Session 1: Tree-Based Methods & Boosting (Part II) (2 hours). Hands-on examples and discussion of hyperparameter tuning. Gaussian Process Regression strategies. 13:15 – 14:15 | Lunch (1 hour) 14:15 – 15:15 | Session 2: Symbolic Regression Overview of symbolic regression techniques for deriving interpretable models (1 hours).

In presenza

Presentazione orale

P.D.2-2 - Settembre

18 - 20 September 2025