NTNU IME IDI

TDT4173 Machine Learning and Case Based Reasoning

 
Menu
  • Main Page
  • Curriculum
  • Lectures
  • Previous Exams
  • Exercises

  • Autumn, 2013

    This page is obsolete. All information is available on ItsLearning.

    Welcome to the IDI course in automated learning from observations, data mining, experience capture, automated model generation, analogy learning, case-based reasoning, hypothesis evaluation, ... The course gives an overview of methods that enable computers to automatically, or semi-automatically, learn problem solving performance from observations in its environment. The first part of the course describes foundational principles of machine learning, combined with studies of some well-known learning methods. The second part is dedicated to learning by retaining and reusing concrete experiences, i.e. analogical and case-based reasoning and learning.

    The general part is covered by the text book Machine Learning, by Tom Mitchell, and a couple of additional papers. The book is from 1997, but it is still the most widely used text book in Machine Learning world wide. It should be available in Tapir in due time before the start. It is highly recommended to read through the first two chapters before the first lecture!

    The case-based reasoning part is covered by a set of papers. We will also use other material (apers etc.) that will be distributed through the website.

    Please note that the course number a few years ago changed from the earlier IT3704. This reflects that the course is intended both for science (informatikk) and for technology (datateknikk) students, and the change does not imply a change in the course structure or focus. You may want to take a look at the course contents of last year. This year will follow the same general line, although a few modifications can be expected due to recent development in the machine learning field.

    TDT4173 is a graduate level course, intended for Ph.D. and M.Sc. (doktorgrad and hovedfag) students of both the science and the technology study lines (teknologifag og realfag/almenvitenskapelige fag). It is assumed that you have a basic knowledge of artificial intelligence, for example corresponding to the course TDT4171 Metoder i kunstig intelligens; including a general idea of classical inductive and deductive learning. Essential issues will be briefly repeated, and if you should detect some holes in your background during the course, you will be pointed to literature that will help you fill them.

    There will be 4 exercises which together with an oral presentation will constitute 20% of the final mark on a pass/fail basis, together with 80% from the written exam. You are not allowed to bring any documents or other help tools (except a simple calculator) to the exam.

    There is also a formal course description available.

    Important information

    • 2012-10-31: We have had a reference-group meeting. Take a look at the the minutes, and send any comments you have to Helge or the group members.
    • 2012-10-25: The last assignment is out. Delivery deadline is Nov 15th. Use It's Learning.

    Language

    All material is in English for the benefit of international students. The course will be taught in English if attended by foreign language speakers, otherwise in Norwegian.

    Exam

    The exam is held on December 6th, 2012. 4 hours.

    Assistance

    • Help with assignments can be obtained Wednesdays 1215 - 1400, when Lars Bungum helps you out in his office (Room 359, IT Vest).
    • If you need to speak to Lars outside the scheduled hours, please contact him by email to arrange a meeting.

    Staff
     
  • Helge Langseth
  • Lecturer/Responsible for the course IT-Vest 310
     
  • Lars Bungum
  • Teaching Assistant IT-Vest 359

    To get in contact with the teacher, use email helgel@idi..., if you have questions/comments regarding the exercises, contact the assistants at larsbun@idi.... Do not hesitate to contact us, feedback from you is the best way for us to improve the course!

    Student Course Evaluation Group (Referansegruppe)

    Minutes from meetings:


    This year's group members:
    Name E-mail
     
  • Jasmin Spago
  • jasmins@stud.ntnu...
     
  • Martin Belgau Ellefsroed
  • ellefsro@stud.ntnu...




    Editor: Head of department : Maria Letizia Jaccheri   Contact address: Helge Langseth   Page updated: