AI Guild Series - Session 3 - Dimensionality Reduction for Data Visualization

Dimensionality Reduction for Data Visualization: PCA vs LDA vs t-sne

When dealing with huge volumes of data, a problem naturally arises.
How do you whittle down a dataset of hundreds or even thousands of variables into an optimal model?
How do you visualize data through countless dimensions? Fortunately, a series of techniques called dimensionality reduction aims to help alleviate these issues. It estimates how informative each variable is and, if needed, skim it off the dataset.

There is no best technique for dimensionality reduction. Instead, the best approach is to use systematic controlled experiments to discover what dimensionality reduction techniques, when paired with your model of choice, result in the best performance on your dataset.

[prasadbhandarkar]

Hey Aishwarya is it possible for you to upload this ipynb file for the reference .