I’ve been studying linear algebra and I decided I would see what it takes to code a multiple regression model from scratch to deepen my understanding of some of the things I’ve been learning. I couldn’t really find any Julia examples so I’m posting this here in the event that others (as I did) searc…

[image] mthelm85: β = [inv(X'X)X'Y] Don’t do this. This is the textbook “normal equations” formulation of linear least-squares problems, and is often a bad idea in a practical numerical context because it squares the condition number (basically: it doubles the number of digits that you lose…

Here’s another approach (because I’m a glutton for punishment :rofl:): using Random using SymPy X = 1:4 Y = rand(collect(-1:0.1:1), 4) .+ collect(1:2:8) # We want to find values for β₁ and β₂ ŷ(β₁,β₂,x) = β₁ + β₂*x @vars β₁ β₂ x y real=true # A function to calculate the residual at a given poi…

Matrix formulation of multiple regression

Specific Domains Machine Learning

dmbates January 14, 2020, 7:43pm 3

Topic		Replies	Views
Efficient way of doing linear regression Performance regression	44	21474	February 7, 2022
Structured multivariate regression Statistics question , linear-regression	2	182	November 21, 2025
Linear Regression step by step for newcomers Machine Learning question	7	6072	November 2, 2022
GLM is slow on large datasets. Using OnlineStats for regressions? MixedModels? Performance glm	25	5414	November 26, 2018
Strange error when performing linear regression General Usage	3	357	December 18, 2021

Matrix formulation of multiple regression

Related topics