From the course: Complete Guide to AI and Data Science for SQL Developers: From Beginner to Advanced
Unlock the full course today
Join today to access over 23,200 courses taught by industry experts.
Model performance comparison on train and test data - SQL Tutorial
From the course: Complete Guide to AI and Data Science for SQL Developers: From Beginner to Advanced
Model performance comparison on train and test data
- [Instructor] Welcome to step 19, where you assess how well your model performs on both the training and test data sets. Now, why are you doing this? Think of it this way. After building a high performance sports car, you'd want to test it out on various terrains and conditions to make sure it's running smoothly. Similarly, you want to evaluate your model's performance in different scenarios to ensure it's reliable. In this step, you'll use three important metrics to gauge your model's performance. Let's introduce these metrics before diving into the results. First, the RMSE, or root mean square error. This is similar to calculating the typical gap between your predicted home values and their actual value. In other words, it tells you how far off, on average, your predictions are from real home values. A lower RMSE is better, because it means your model's predictions are closer to reality. MAE, or mean absolute error.…
Practice while you learn with exercise files
Download the files the instructor uses to teach the course. Follow along and learn by watching, listening and practicing.
Contents
-
-
-
-
-
-
-
-
-
-
-
-
-
(Locked)
Creating the linear regression model and model summary: Part 19m 33s
-
(Locked)
Creating the linear regression model and model summary: Part 27m 16s
-
(Locked)
Creating the linear regression model and model summary: Part 35m 33s
-
(Locked)
Dropping insignificant variables and re-creating the model7m 57s
-
(Locked)
Checking assumptions for linear regression3m 18s
-
(Locked)
Assumption 1: Checking for mean residuals2m 47s
-
(Locked)
Assumption 2: Checking homoscedasticity3m 13s
-
(Locked)
Assumption 3: Checking linearity2m 12s
-
(Locked)
Assumption 4: Checking normality of error terms3m 24s
-
(Locked)
Q-Q plot for checking the normality of error terms3m 14s
-
(Locked)
Model performance comparison on train and test data6m 7s
-
(Locked)
Applying cross-validation and evaluation4m 40s
-
(Locked)
Challenge: Model building48s
-
(Locked)
Solution: Model building1m 16s
-
(Locked)
-
-
-