ICP6 - narhirep/Python-Deep-Learning GitHub Wiki

Welcome to In Class Programming 6 :

Description: This assignment is about getting to know clustering algorithms, data normalization and principle component analysis.

Objective: To solve the given problems with the help of KMeans clustering algorithm, then to standardize the data and to convert the high dimensionality to low dimensionality by using PCA.

Implementation:

Below are the questions for ICP6.

For first question we need to apply K means clustering in the data set provided So we need to remove any null values by the use of mean value. So we are using the elbow method to find a good number of clusters with the KMeans algorithm. Also we are calculating the silhouette score for the clustering shown below.

Here is the elbow graph for question 1.

For second question we are trying scaling feature and then apply KMeans on the scaled features and we got below output which is lesser than first method.

Here is the elbow graph for question 2.

For third question we are applying PCA on the same given dataset which we used above. So after applying kMeans algorithm on the PCA we see that model is performing better than second one but not as good as first one.

Here is the elbow graph for question 3.

Video Link: ICP6

Conclusion: In ICP6 assignment I learnt about KMeans clustering algorithm, data normalization and how to convert the high dimensionality to low dimensionality by using PCA.