Visual_Data_Analytics_Basics

Performing basic visual data analysis using Python (Server side) and D3.js(client side)

Languages and frameworks used

• Languages:
  • Python
  • Javascript
  • HTML
  • CSS
• Libraries:
  • Flask (https://pypi.org/project/Flask/)
  • D3.js (https://d3js.org/)

Clustering of Data

• Experimentation with two types of sampling: Random and Stratified (k-means with optimized k using elbow method)

## Dimensionality Reduction - Comparison of effects of dimensionality reduction on the original and the reduced data samples obtained - Performed PCA on the data samples to produce scree plot and to compare them - Obtained the intrinsic dimensionality of the data sample - Obtained top three attributes with the highest PCA loadings - Performed MDS on data samples (using both Euclidean and Correlation distance) to compare them

Scatterplot Visualisations

• 2D Scatterplot visualization for top two PCA vectors for all the data samples
• 2D Scatterplot visualization for the MDS data (Euclidean and Correlation)
• Scatterplot Matrix visualization for the top three PCA loaded attributes

Scatterplot: PCA

Scatterplot: MDS

Scatterplot: ScatterplotMatrix

DEMO VIDEO LINK (YouTube):

https://youtu.be/U9lX7lheJXA