Hide menu

Artificial Intelligence

DF22200, 2011VT

Status Archive
School National Graduate School in Computer Science (CUGS)
Division AIICS
Owner Erik Sandewall

New version of course from spring 2011
Old course code DF00600

  Log in  




Course plan

Lectures

24 hours

Recommended for

Doctoral students in computer science.

The course was last given

2007, but with another course plan

Goals

The student should learn to understand and to use basic formal and practical techniques in artificial intelligence

Prerequisites

Solid competence in computer programming Introductory course in artificial intelligence Previous knowledge of first-order logic is recommended, but can also be obtained using additional study within the course

Contents

The contents are structured as core contents and additional contents.

Core Contents
-------------

Basic concepts and formal languages in knowledge representation

Decision trees and causal nets

Description logic

Defeasible inheritance

Reasoning about actions and action planning:
Characterizing actions in terms of preconditions and effects
Use of preconditions in autonomous agents
Explicit-time logics and the situation calculus
Prediction; the frame problem and related issues
Action planning methods

Nonmonotonic reasoning. Entailment using preferred models, default logic, circumscription

Search techniques, the A* algorithm, SAT methods

Computational paradigms in artificial intelligence: logic programming and logic-programming-based approaches, answer-set programming, production systems, rule-based systems

Software architectures for intelligent autonomous agents: SOAR, BDI architectures, Hierarchical Task Networks, three-level architectures

Additional Contents
-------------------

Some asepcts of computer learning: Reinforcement learning, inductive learning, learning of decision trees, classification, PAC learning

Qualitative reasoning

Software issues in robotics

Model-based diagnosis and execution monitoring

Organization

Lectures

The course is organized as three course weeks, with several lectures during each of those weeks, and several weeks of independent study between the course weeks. It is assumed that the students can acquire major parts of the course contents by studying the lecture notes.

Literature

Lecture notes as specified on the course webpage

Lecturers

Erik Sandewall

Examiner

Erik Sandewall

Examination

Completion of exercises ("labs")
A written exam

Credit

6 hp

Comments

Course webpage


Page responsible: Director of Graduate Studies