Tools of Artificial intelligence / Kunstig intelligens værktøjer

The Maersk Mc-Kinney Moller Institute, Odense
Teaching activity id: RMAI2-U1.
Teaching language: English.ECTS / weighting: 5 ECTS / 0.083 full-time equivalent.
Period: Spring 2016.Approved: 14-05-14.
Offered in: Odense.

Subject director:
Ole Dolriis, The Maersk Mc-Kinney Moller Institute.

Prerequisites:
Basic object oriented programming knowledge is recommended.

Learning outcomes:
Knowledge:

+ understanding of basic machine learning techniques:  neural
  networks, genetic algorithms and reinforcement learning

+ understanding of representational techniques appropriate to the
  above learning methods

+ understanding of the experimental challenges and demands of the
  machine learning approaches referred to above

Skills:

+ able to implement, debug and deploy the AI techniques taught in new
  situations

+ able to devise suitable representations of data for chosen machine
  learning techniques

+ able to test, evaluate and document the performance of chosen
  machine learning techniques using suitable correct methodologies

+ able to write a straightforward experimental scientific paper
  documenting a comparison experiment

Competences:

+ can identify robotic problems where machine learning techniques
  could be applied

+ can select appropriate techniques from the toolbox of possibilities

+ able to characterise a new AI technique in terms of scope and
  type (unsupervised, semi-supervised or supervised)

+ can evaluate reported applications of machine learning techniques
  in terms of results and methodology



Lessons:
48 lessons

Form of instruction:
Lectures and exercises.

Evaluation
Individual written report based on project and evaluated according to the Danish 7-point grading scale with external co-examiner.

Programmes:
Master of Science in Engineering (Robot Systems)
2. semester, mandatory. Offered in: Odense