Implementing a Linear Regression from Scratch - MSE & Gradient Descent - Tutorial - 2025

Описание: Implementing a Linear Regression from Scratch VIA Applying the Gradient Descent on our Loss function (which will be the Mean Squared Error).

This video covers everything, from the concept of Residual, Loss functions, the MSE, The gradient Descent Simplified, The Gradient Descent on a Bivariate Function, Partial Derivatives and Python implementation (in Google Collab)

Resources:
Github Repo (everything is here): https://github.com/GDSimpson3/LinearR...
Dataset: https://raw.githubusercontent.com/GDS...
Notebook: https://github.com/GDSimpson3/LinearR...
Board: https://github.com/GDSimpson3/LinearR...

Tools used:
Google Colab
Microsoft Whiteboard
OBS

AI Community
  / discord  

CHAPTERS
0:00 Introduction
0:47 Getting our Dataset
5:00 Visualise the Data
8:15 Regression Line Concept
10:20 Loss Functions
10:30 Residual
12:16 Mean Squared Error
20:16 The Gradient Descent on a Simple Quadratic
29:56 The Gradient Descent on our Bi Variate Function
32:19 Partial Derivatives Explanation
36:41 Differentiating a Summation
38:12 Partial Derivatives of our MSE
41:15 Gradient Descent with Partial Derivatives
46:55 Implement in Code
47:05 Drawing a line
50:26 Implementing the MSE
57:22 Partial Derivatives implementation
01:00:09 Gradient Descent Implementation
01:05:12 Visualise the MSE
01:07:35 Pitfalls of the Gradient Descent
01:08:56 Extra

TAGS
#linearregression #regression #AI #ML #2025 #machinelearning #beginner #residual #statistics #partialderivatives #gradientdescent #firstprinciples #implementation #python #jupyterlab #gradients #iteration #convergence #tutorial #alpha #learningrate