SPPU · Final Year Computer Engineering · 2019 Course (Revised)
Pattern Recognition
410253A · Elective VI · Semester VIII · 3 credits · In-semester 30 + end-semester 70
One of the Elective VI options.
Unit 1: Pattern Recognition
7 hoursApplications, pattern recognition systems, the design cycle, learning and adaptation, representing patterns and classes, feature extraction, approaches.
Marks: 15 in the in-semester exam
- Pattern recognition systems and the design cycle
- Pattern representation and feature extraction
- Pattern recognition approaches
Unit 2: Error Estimation & Decision Theory
7 hoursError estimation; Euclidean, Manhattan, cosine and Mahalanobis distances and distance-based classifiers; feature selection by hypothesis testing, ROC curves; Bayesian decision theory, minimum-error-rate classification, discriminant functions; maximum likelihood and Bayesian estimation, dimensionality; Parzen windows, nearest neighbour, fuzzy classification.
Marks: 15 in the in-semester exam
- Distance measures and distance-based classifiers
- Error estimation and ROC curves
- Bayesian decision theory
- Discriminant functions
- Maximum likelihood and Bayesian parameter estimation
- Non-parametric methods: Parzen windows and nearest neighbour
Unit 3: Structural Pattern Recognition
6 hoursDecision trees and random forests; formal grammars, string patterns, syntactic recognition and parsing, stochastic grammars; graph representations; Boltzmann learning.
Marks: 17–18 in the end-semester exam
- Tree classifiers: decision trees and random forests
- Syntactic pattern recognition with grammars
- Graph-based structural representation
- Boltzmann learning
Unit 4: Clustering
8 hoursHierarchical and agglomerative clustering, single, complete and average linkage, Ward's method; K-means; MST-based clustering; clustering by simulated annealing.
Marks: 17–18 in the end-semester exam
- Hierarchical clustering and linkage methods
- K-means clustering
- Graph-based clustering
- Clustering with simulated annealing
Unit 5: Template Matching and Unsupervised Learning
7 hoursBellman's optimality principle and dynamic programming, edit distance, dynamic time warping, correlation measures, deformable templates.
Marks: 17–18 in the end-semester exam
- Template matching and dynamic programming
- Edit distance
- Dynamic time warping
- Correlation and deformable templates
Unit 6: Fuzzy Logic and Pattern Recognition
7 hoursFuzzy logic and fuzzy classifiers; classification with genetic algorithms; object, face, iris, fingerprint and 3D recognition.
Marks: 17–18 in the end-semester exam
- Fuzzy pattern classifiers
- Pattern classification with genetic algorithms
- Biometric recognition: face, iris and fingerprint
Unit contents are summarised from the official SPPU syllabus . AiCanCode.org is not affiliated with Savitribai Phule Pune University.