PetePrattis / user-clusters-and-k-means-fold-for-classifier-evaluation

A Matlab script that applies the basic sequential clustering to evaluate the number of user groups by using the hierarchical clustering and k-means algorithms. Using the k-means fold the classifiers that are a neural network and the other least squares to evaluate them.

A Matlab Exercise / Project

This is a Matlab project from my early days as a Computer Science student

This programm was created for the fifth semester class Pattern Recognision and it is the final project necessary to pass the class

Description of project

A Matlab script that applies the basic sequential clustering to evaluate the number of user groups by using the hierarchical clustering and k-means algorithms. Using the k-means fold the classifiers that are a neural network and the other least squares to evaluate them.

Implementation of project

1. Apply the basic sequential schema to estimate the number of user groups according to their preferences.
2. Based on the estimation of Step 1, apply the k-means algorithm and the hierarchical clustering algorithm.
3. Using the 5-fold format provided, design, implement, and evaluate two classifiers, which solve the following problem: if a user and a movie is given, the classifier decides whether the user saw the movie (class 1) or not (class 2) . One classifier will be a neural network and the other a least squares.

About this project

• The MovieLens dataset used for 100k ratings https://grouplens.org/datasets/movielens/100k/
• This program was written in Matlab IDE
• This repository was created to show the variety of the work I did and experience I gained as a student