Intelligent Sensorial Processing

Objectives

To know:

  • Automatic classifiers design
  • Automatic classifiers performance evaluation
  • Applications of decision support systems

To do:

  • Development of automatic classifiers

Integration of decision support systems

General characterization

Code

10964

Credits

6.0

Responsible teacher

José Manuel Matos Ribeiro da Fonseca

Hours

Weekly - 4

Total - 56

Teaching language

Português

Prerequisites

None

Bibliography

Leo Breiman, Jerome Friedman, Charles J. Stone, R.A. Olshen. Classification and Regression Trees. Chapman and Hall, New York.

Quinlan, Ross J. (1993). C4.5: Programs for Machine Learning. San Mateo, California. Morgan Kaufmann.

Weiss S. M., Kulikowski C. A. (1991) Computer systems that learn. San Mateo, California. Morgan Kaufmann.

Fonseca, José M. (1994). Indução de Árvores de Decisão. Tese de Mestrado

 

Teaching method

Theorectical and practical classes

Evaluation method

One evaluation test and a project with presentation. Final exam if necessary.

Subject matter

• Introduction to automatic classifiers design
• Estimation of a classifier quality
• Notion of error and cost
• Methods for estimating the error of classifiers.
• Nearest-neighbor classifiers.
• Decision trees.
• Methods for the partition of discrete and continuous characteristics
• Criteria for selection of attributes
• Methods for limiting the growth of decision trees
• Decision trees forests
• Neural networks.
• The perceptron
• Back-propagation
• Multi-layer neural networks

• Classifiers based on genetic algorithms.
• Fuzzy classifiers
• Principles of fuzzy controllers
• Fuzzy inference
• Fuzzy Algebra
• Biometric techniques
• Identification systems based on biometrics: iris recognition, retina, hand, fingerprint, face, etc.

Programs

Programs where the course is taught: