Pattern Recognition

University course KPI IPT 2019y

All needed tasks described in all-labs.pdf

Solutions described in solved-tasks.pdf

lab1 - Bayesian Strategy for Binary and Delta Loss Function

Image Denoising (Bernoulli noise)

lab2.1 - Analysis of Nonbayesian Strategy

lab2.2 - Implementation of Effective algorithm for calculation cumulative sums of 1D, 2D and 3D arrays

lab3 - Hidden Markov Model (HMM)

Divisibility sum of digits

lab4 - DP algorithm for chain-structured graphical models

Binocular stereo vision

NOTE: Source of images - http://vision.middlebury.edu/stereo/data/

lab5 - EM Algorithm for Binary Clustering

Binary image clustering (mnist dataset)

lab6 - Modification of a Perceptron Algorithm

Modification of Perceptron Algorithm where hyperplane is a circle (modified scalar product)