Skip to content

42028 - Intelligent Decision Support Systems (SISPD) [UPC]


Type: S3 Course
Semester: Fall
ECTS: 6
Teaching Points: 15
Offer: Annual
Responsible Unit: CS-UPC
Responsible: Miquel Sànchez i Marrè
Language: English
Requirements:

GOALS

The course focuses on analysing the inherent complexity of most real-world systems or domains, and on the need to use decision-making support tools.

Decision theory and several decision-making models will be studied. Some computational decision support systems will be reviewed, and finally, a proposal of a new paradigm within the artificial intelligence field will be made: Intelligent Decision Support Systems (IDSS). Architecture of IDSSs, requirements for their construction, their implementation through a computer, and their limitations and advantages will be discussed.

Knowledge discovery process from data in IDSSs will be deeply studied, and especially, its inherent multidisciplinary nature will be outlined, emphasising the interaction between artificial intelligence techniques and statistical techniques. This later interaction provides IDSSs with different but complementary knowledge models to be used within the scope of a IDSS to solve different kind of real-world domains.

Validation techniques for knowledge models development and available tools will be shown within the course too.

Finally, several real case studies will be described and analysed, especially within the environmental systems field. Some tools will be used to develop a small-medium project related with IDSSs development and with real data.


CONTENTS

1. Introduction
  • Complexity of real-world systems or domains
  • Need of decision-making support tools
2. Decisions
  • Decision Theory
  • Decision process modelling
3. Decision Support Systems
  • Traditional tools
  • Computational tools
4. Intelligent Decision Support Systems (IDSS)
  • IDSS Architecture
  • Requirements, features and drawbacks of IDSSs
  • Analysis and design of an IDSS
  • Implementation of an IDSS
  • IDSS Validation
5. Knowledge Discovery in an IDSS: from data to models
  • Introduction
  • Knowledge Discovery and Data Mining
  • Artificial Intelligence and Ststistics
  • Data structure
  • Data filtering and Data preparation
  • Knowledge models and Data Mining techniques
  • Descriptive Models
  • Variable Association Models
  • Discriminant Models
  • Predictive Models
6. Model Validation
  • Graphical tools
  • Statistical methods for hypothesis verification
7. Software tools for IDSS development:
  • GESCONDA
  • KLASS
  • WEKA
  • Other tools: R system
8. IDSS Application examples and real-world case studies