Vectorized Data and Principal Component Analysis

In machine learning, features and beddings are sometimes called vectors. These embeddings compress the input data in a different representation. It's like if you combine the different spectral bands and the different dates in a time series to get reduced representation which contains meaningful information. Generic embeddings go beyond that, are able to take into account other correlations in the data,. Other dependencies, and more complex information. But exactly, this is a very, very good analogy.

Transcript

Play full episode

The AI-powered Podcast Player

Save insights by tapping your headphones, chat with episodes, discover the best highlights - and more!

Get the app