https://calendly.com/compuflair/data-science-consultation
Schedule a 1 Hour Consulting session to
(1) answer your personalized questions, (2) help get your data science projects done, (3) draw a roadmap for your data science career.
https://compu-flair.com/solution
Do you need to derive insight from your data? Fill out the form on our solution page, explain what you need, and we will get back to you with an offer.
—————–
In this video, Dr. Ardavan (Ahmad) Borzou will describe the concept of entropy and relative entropy in physics, statistics, and information theory and how it can be used to use data to estimate machine learning unknown parameters.
After discussing the concepts, at the end of the video, he will show how to use Python and its libraries to run the implementations of the presented methods to uncover linear regression parameters from a sample data spreadsheet.
Chapters:
00:00 – Introduction
00:41 – Summary of Past Videos
02:28 – Surprise, Action, Free Energy
02:55 – Uncertainty VS Randomness
04:28 – Free Energy Minimum
04:51 – Surprise
05:43 – Entropy
08:26 – Probability Estimation
08:54 – Relative Entropy
10:17 – Add Data
11:42 – Minimal Information Loss
12:20 – Minimization to Maximization Trick
13:11 – Likelihood
14:35 – Analytic Solution to Linear Regression
16:00 – Residual Sum of Squares
17:45 – Implementation in Python
19:36 – Source Code & AI Assistant
20:38 – Coaching or Consulting Services
source