Intelligent Systems
Core concepts of intelligent systems: knowledge representation, reasoning, search algorithms, and agent architectures.
Workgroups/seminars are not mandatory, but attending 6 out of 7 sessions earns a 0.3 bonus point on your final grade (if you pass the course). Practical sessions for assignment help are optional and require sign-up on Canvas.
Learning objectives
Knowledge and understanding: at the end of the course, the students will be familiar with basic knowledge of some of the core aspects of AI: state-space representations, search, adversarial search, logic, automated reasoning, reasoning with uncertainty and vagueness and machine learning. Applying knowledge and understanding: students will be able to implement basic (adversarial) search algorithms, as well as knowledge based and adaptive methods to build Intelligent Agents. Making judgements: the students will have a basic understanding of the ethical and societal implications of the developments in AI. Communication skills: students will be able to write a scientific reports about an original research question in a group of students. Learning skills: students will be trained in acquiring a set of complex AI related topics in a restricted period of time, come up with an original research question and perform the necessary (empirical) research.
The Intelligents Systems course focuses on the development and scientific analysis of methods for rational agents. The course is assessed through two individual partial exams (70% total) and 11 practical assignments (30%). The first partial exam covers lectures 1-7 (Introduction, Rational Agents, State-Space Representations, Uninformed & Informed Search, Adversarial Search, and Logical Agents). The second partial exam covers lectures 8-13 (Vagueness, Uncertainty & Bayesian Networks, and Machine Learning). Both exams must be passed with a minimum grade of 5, and the average of the two must be at least 5.5. For the practical assignments, you must pass at least 7 out of the 11 assignments given in weeks 2-7. The grade scales linearly from 6.5 (7 passed) to 10 (11 passed). You must pass both the exams component and the practical component to pass the course. Additionally, you can earn a 0.3 bonus point on your final grade if you participate in at least 6 out of the 7 working groups. Grades between 5.0 and 6.0 are rounded to the nearest whole number (meaning a 5.4 is a fail, and 5.5 is a 6), while all other grades are rounded to the nearest 0.5.
Assessment
The grade will be determined via two (digital) exams (35%+35%), and a number of practical assignments (30% in total). All three components have to been completed successfully to pass the course. There will be a (single) resit exam, combining both partial results, but no resits for the practical assignments.
Teaching methods
2 lectures of 2 hours per week. Tutorials to get started with the programming exercises in the first weeks. Working groups to practice the theoretical material (2 hours). Practical groups to apply the acquired knowledge (flexible, on demand)
Literature
Chapters from "AI a Modern Approach", Russell and Norwig, provided as a part of a reader.