DBATU · Final Year B.Tech Information Technology · Non-NEP
Pattern Recognition
BTITE703A · Elective I · Semester VII · 3 credits · Continuous assessment 20 + mid-semester 20 + end-semester 60
One of the Elective I options.
Unit 1: Introduction
Machine perception; pattern recognition systems: sensing, segmentation, feature extraction, classification and post-processing; the design cycle; supervised, unsupervised and reinforcement learning; applications.
Examined in the 60-mark end-semester paper, and in the mid-semester exam if it is taught by then
- Pattern recognition systems and the design cycle
- Feature extraction and classification
- Supervised, unsupervised and reinforcement learning in pattern recognition
Unit 2: Statistical Pattern Recognition
Bayesian decision theory and classifiers; the normal density and discriminant functions; linear discriminant functions; decision boundaries and separability; the perceptron and support vector machines.
Examined in the 60-mark end-semester paper, and in the mid-semester exam if it is taught by then
- Bayesian decision theory
- Normal density and discriminant functions
- Linear discriminant functions and the perceptron
- Support vector machines
Unit 3: Parameter Estimation and Neural Networks
Maximum-likelihood and Bayesian parameter estimation; PCA and Fisher's linear discriminant; expectation–maximisation; the multilayer perceptron; a brief look at deep, convolutional and recurrent networks.
Examined in the 60-mark end-semester paper, and in the mid-semester exam if it is taught by then
- Maximum-likelihood and Bayesian parameter estimation
- PCA and Fisher linear discriminant
- Expectation-maximisation
- Neural networks for pattern recognition
Unit 4: Nonparametric Techniques
Density estimation; Parzen windows; k-nearest-neighbour estimation and the nearest-neighbour rule; fuzzy classification; decision trees (CART, C4.5, ID3) and random forests.
Examined in the 60-mark end-semester paper, and in the mid-semester exam if it is taught by then
- Parzen windows and density estimation
- K-nearest neighbour rule
- Fuzzy classification
- Decision trees: CART, ID3 and C4.5
- Random forests
Unit 5: Unsupervised Learning and Clustering
Criterion functions for clustering; k-means, hierarchical clustering, k-medoids and DBSCAN; cluster validation; feature selection and extraction; classifier ensembles: bagging and boosting (AdaBoost).
Examined in the 60-mark end-semester paper, and in the mid-semester exam if it is taught by then
- Clustering criterion functions
- K-means, K-medoids and hierarchical clustering
- DBSCAN and cluster validation
- Feature selection and extraction
- Bagging and boosting: AdaBoost
Unit contents are summarised from the official DBATU syllabus . AiCanCode.org is not affiliated with Dr. Babasaheb Ambedkar Technological University (DBATU).