Singular Value Decomposition and LoRa

In this post we:

Understanding the Universal Approximation Theorem for single variable real valued functions

[Note: All of this code used in this post can be found at this colab link, but with a slightly different exposition.]

Fitting linear regression

It turns out that all three of the following problems use exactly the same ideas, but with increasing levels of mathematical sophistication.

Understanding Binary Cross-entropy

Maximum Likelyhood Estimation and Binary Cross-entropy

Fitting logistic regression

Explaining the Model