This project contains 2 parts.
- Part 1 consists of implementing Principal Component Analysis algorithm from scratch without using any libraries for PCA.
The goal is to project high dimensional data to 2 dimensions, and visualize the results. - Part 2 consists of performing Association analysis by implementing the Apriori algorithm from scratch without using
any apriori libraries.
The goal is to generate Association rules and then mine those rules based on given constraints. The format of these constraints are mentioned in the template.pdf
Given a high dimensional data, we projected this data to 2 dimensions, by implementing following techniques -
- Principal component analysis - PCA (implemented from scratch)
- Singular Value Decomposition - SVD
- t-SNE
Based on the given gene data, we implemented the Apriori Algorithm from scratch and generated frequent itemsets
for given support.
Generated Association rules based on the generated frequent itemsets for given confidence.
Mined association rules based on the constraints provided in template.