Joint Call for proposals – AI Projects (BSc/MSc student groups)

• Takes place in Q2, 11th of November 2024 to the 31st of January 2025
• Prior knowledge: Python, basic probability theory, basic linear algebra, machine learning (theory and application with scikit-learn), deep learning (Pytorch), natural language processing (including transformers).
• Project coordinators: Tom Viering, Elena Congeduti, Przemysław Pawełczak 

The minor programme Engineering with AI provides foundational knowledge at the BSc level for students to understand what AI and machine learning is. The minor takes place in Q1-Q2. In the first quarter, students learn about Python, machine learning and ethics. 

In the second quarter, the students first learn about natural language processing (NLP) and deep learning (DL). In parallel, during the first weeks of Q2, you will have the first meetings with the students to clarify and set the project goals and students will work out a project plan. The students submit this, and this needs to be graded and approved before the Christmas break. After agreeing together with the students on the project plan and requirements, the students work on the project full-time after the Christmas break on the project for 3 weeks (+- 40 h per week). They submit their AI solution (deliverables such as code, results, model weights / AI solution, collected data, etc.) in week 9, and in week 10 they present their AI solution to you.  

The project work requires the application of AI (machine learning, deep learning) in a domain of one of the faculties. Next to supervision of a staff member, postdoc or PhD student, the students are supported by a teaching assistant who can provide help with basic questions. As a supervisor, it is not mandatory to have all AI knowledge – you can function as a domain expert / stakeholder / or “client”.  The course coordinators in this case can also help you write the project proposal to ensure feasibility and sufficient challenge for the students. The student group will showcase their results in a final presentation for the assessment. There is no report for this course. 

• Takes place in Q1, September to November 2024
• Prior knowledge: Python programming; Additional prerequisites could be specified for each project (e.g., knowledge of quantum physics).
• Basic knowledge of AI methods (equivalent of 10EC) will be gained by students during the programme.
• Project coordinators: Arkady Zgonnikov, Ashwin George 

The Fundamentals of AI Programme (FAIP) provides foundational knowledge at MSc level for students to understand what Artificial Intelligence is. This programme is aimed at students without previous knowledge of AI. During the programme the students will obtain a thorough understanding of AI, focusing on core AI methods and topics, including data handling, machine learning, societal impact, strengths and weaknesses of methods, and key application areas. 

This programme runs in Q1. In weeks 1.1 – 1.5 students follow intensive lectures on selected AI topics. During week 1.1, students will also be assigned to project groups based on their preference. In weeks 1.2-4, each group will have initial meetings to introduce themselves to their supervisor and together define exact project requirements. From weeks 1.5-6 students work on the project in groups of 4-5. The project work requires the application of AI in a specific domain. The applied AI techniques, as well as the required entry level of programming and domain knowledge, can vary across projects and domains. The tangible output of the project is a final report, presentation, and optionally the source code shared with the supervisor.

• Takes place in Q1, Takes place in Q1, 2nd of September to the 8th of November 2024
• Prior knowledge: 20 EC on AI related courses (at Bachelor or Master level). Since this is an interdisciplinary team the members of the team can different AI-related experience, e.g., some members may have experience in developing AI solutions, others in applying these solutions to a problem domain, and some others on assessing the impacts.
• Project coordinator: Pradeep Murukannaiah, Jazmin Zatarain Salazar 

Having a solid shared pool of knowledge on AI enables the students embarking on this block to work on an AI challenge at a master level. This project provides students with a unique experience to learn to develop and apply AI to a real-world challenge and assess its impact, while working in an interdisciplinary project team. 

The target group of this course are all TU Delft students with experience in AI. They could work as applied AI developers / researchers in a domain in their professional careers. This project runs in Q1. AI2P aims to provide students with a unique experience to learn, develop and apply AI to a real-world challenge, while working in an interdisciplinary project team. During the project, the team executes AI research, and/or produces AI solutions, that address real-world problems. Students collaborate in a group of 3-5 students with (ideally) diverse backgrounds (including, e.g., Computer Science and Robotics).