4. Algorithmic Decision Theory Lectures
Raymond Bisdorff, Emeritus Professor of Applied Mathematics and Computer Science
Bisdorff © 2013-2022
From 2007 to 2011 the Algorithmic Decision Theory COST Action IC0602, coordinated by Alexis Tsoukiàs, gathered researchers coming from different fields such as Decision Theory, Discrete Mathematics, Theoretical Computer Science and Artificial Intelligence in order to improve decision support in the presence of massive data bases, combinatorial structures, partial and/or uncertain information and distributed, possibly interoperating decision makers.
A positive result a.o. of this COST action was the organisation from 2012 to 2020 of a Semester Course on Algorithmic Decision Theory at the University of Luxembourg in the context of its Master in Information and Computer Science.
Below are gathered 2x2 reduced copies of the presentation slides for 12 Lectures from the Summer Semester 2020.
- L1. General introduction to Algorithmic Decision Theory
Historical notes and acknowledgements
Generic conceptual framework for studying decision aiding processes
Selecting, ranking, rating and clustering problems
- L2. Who wins the election ? Choosing from multiple opinions
On plurality tyranny in uni-nominal elections and other difficulties with simple voting rules
How to aggregate voter’s preferences ?
Voting and complexity issues
- L3. On social consensus rankings
On ranking from different opinions
A typology of ranking rules
Classification of ranking rules
- L4. Evaluation models for measuring and aggregating performances
Rules for aggregating grades
How to aggregate ordinal grades ?
- L5. Solving social compromise decision problems with CBA
What is Cost-Benefit Analysis (CBA) ?
Principles and critical perspective
Applications in public transport problems
- L6. Choosing with multiple commensurable criteria: the Multiple Attribute Value Theory
Measuring the performances of potential decision alternatives
Agregating Costs and Benefits
Theoretical foundations and critical perspective
- L7. Choosing with multiple non-commensurable criteria: The Rubis outranking approach
Comparing alternatives with potentially conflicting criteria
Theoretical foundation of the outranking approach
The Rubis best-choice recommender system
- L8. Generating random outranking digraphs
Random performance generators
Random standard performance tableau
Special models: Cost-Benefit, 3-Objectives or academic performance tableaux
- L9. On rating with multiple performance criteria
How to rate with multiple incommensurable criteria ?
On rating-by-sorting with relative quantiles
Absolute rating-by-ranking with learned quantile norms
- L10. On ranking from bipolar-valued pairwise outranking situations
Ranking with outranking digraphs
- L11. On ranking by first and last choosing
Partial weak ranking by first and last choosing
Useful properties of the Rubis best-choice procedure
A bipolar ranking-by-choosing algorithm
- L12. Ranking big multiple incommensurable criteria performance tableaux
Pre-ranking a q-tiled performance tableau
On sparse outranking digraphs
HPC-ranking of big performance tableaux