Information
| Course Code | EIP206 |
| Semester | 2 |
| Category | Basic |
| ECTS Credits | 8 |
| Eclass |
Professors
Proposed Bibliography
R.O. Duda, P.E. Hart and D. G. Stork, “Pattern Classification”, Wiley, 2nd Edition, 2001.
Η. L. Van Trees, “Detection, Estimation and Modulation Theory”, Wiley, 1971.
Course Description
- Basic principles of classic classification theory (Bayes). Probability ratio as criterion in population distinction. Use of the theory in standard (Gaussian) deviate populations.
- Mahalanobis Distance: Separation of the feature space according to the population statistics and the correlation of the features.
- Correlation of population features: Degree of correlation. Feature quality. Dimensionality of a classification problem. Dimensionality reduction and important dimensions.
- Artificial Neural Networks: Problems that they can solve. Neural networks simple structures.
- Parameters Estimation. Calculation of the distribution of dependent random variables:
- Mathematical Morphology
- Signal detection theory: Basic concepts. Neyman-Pearson criterion. Constant false alarm rate (CFAR) detector.
- Information synthesis, into simple data, into features and into decisions.
- Remote sensing
- Examples of remoting-sensing information synthesis.
- Decision synthesis.
- Oversampling – Noise shaping – ΣΔ coders
