Data Magic: How Dimensionality Reduction Boosts Analytics
Dimensionality Reduction: Unraveling its Superpowers
Top 5 data analyses using dimensionality reduction
Did you know that dimensionality reduction techniques can do more than reduce dimensions? In this story, I will show you the top 5 data analyses which can be done using techniques such as Principal Component Analysis (PCA) and t-distributed Stochastic Neighbor Embedding (TSNE).
In order to illustrate the story, I will take data related to video game sales. Sample data is shown here
The data has game details as well as sales figures. Let us now see different analyses which are possible using dimensionality reduction.
2D Visualisation
The video game data has 9 dimensions, and it is impossible as humans to visualize such high dimensional data. One of the apparent use of dimensionality reduction is visualizing the data. Shown below is the result of the most common dimensionality reduction technique, PCA, which helps in projecting 9 dimensions to 2 dimensions.
