Knowledge Based Systems

The theoretical content of the course will be delivered through lectures, supplemented with case studies, whenever possible.
Students will be encouraged to apply the skills acquired through practical activities.

This discipline introduces the principles of Knowledge Based Systems and the Theory of Knowledge Discovery in databases, presenting their concepts, phases, key tasks and algorithms involved, and their possible application in real problems, such as financial planning, production management, diagnosis, prediction or detection of fraud.
The Knowledge Discovery (KD) is a subarea of artificial intelligence that comes to discovering patterns and relationships in large data and text. The knowledge discovery involves technologies such as data mining, mining the text rule induction, group data, and other related techniques. Developed and commercial tools of KD are applied to real problems of financial planning, production management, diagnosis, prediction, fraud detection, etc.

1. Knowledge Based Systems
1.1 Concepts and definitions
1.2 General structure of a Knowledge Based System
1.3 Process development of knowledge based systems
2. Data Mining and Knowledge Discovery in Databases
2.1 Data Mining
2.2 Definition of Data Mining
2.3 Areas related to Data Mining
2.4 Approaches to Data Mining
2.5 Objectives of Data Mining
2.6 Knowledge discovery in databases
3. Methodologies and Specifications
3.1 CRISPDM methodology
3.2 SEMMA methodology
3.3 PMML Specification
3.4 Conclusion
4. Models and Techniques
4.1 Decision Trees
4.2 Artificial neural networks
4.3 Genetic Algorithms
4.4 Induction of rules
4.5 Fuzzy sets
4.6 Rough Sets
4.7 Bayesian Networks
4.8 Classification Systems
4.9 Algorithms evaluation
4.10 Sampling
11.4 Evaluation of models
4:12 Confusion matrix
4:13 Error cost
4.14 ROC curve
4:15 Regression
5. Tools and Technologies
6. Areas of Application

NAO