/AMLT

Course project for Advanced Machine Learning Techniques (AMLT) at UPC

Primary LanguageJupyter Notebook

AMLT Practical Project 2016-17

This repository contains the analysis of two topics in Machine Learning; Distance Based Outlier Detection, and Clustering.

Distance Based Outlier Detection

Among Distance Based Outlier Detection Algorithms, two of them; k-th neares neighbor as known as dee-kay-en, and one-time sampling have been implemented in the following files:

  • kth_NN.py
  • sampling.py

Clustering Algorithm

K-Means and Bisecting K-Means have been implemented in the following two files:

  • kmeans_numpy.py
  • bikmeans_numpy.py

Ipython Notebooks:

An Ipython notebook named "Experiment One - Outlier Detection" has been provided that uses, analyzes and reports the outlier detection algorithm.

Another Ipython notebook named "Experiment Two - Clustering Algorithms" has been provided that uses, analyzes and reports the clustering algorithms.

PDF Files:

Some PDF files that were studied have been attached along with a final report about the research.