On this weekly Office Hours for Oracle Machine Learning on Autonomous Database, Jie Liu, Data Scientist for Oracle Machine Learning, covered the best practices when using Machine Learning Regression metrics.

The goal of a regression task is to build models based on features to predict a target quantity, that is, a numeric value. After a regression model is applied to a test set, the next step is to evaluate the model performance by checking the error between the regression output and the true value. A certain set of metrics is often used to evaluate the regression model such as mean squared error (MSE), root mean squared error (RMSE), mean absolute error (MAE) and R squared, etc. What are the differences among those metrics? How does one choose a metric correctly? How does the metric translate back to the business world?

In this Session, Jie went through various regression metrics and addressed the questions above. Moreover, he showed how to compute regression metrics in a scalable way using OML4Py.

The Oracle Machine Learning product family supports data scientists, analysts, developers, and IT to achieve data science project goals faster while taking full advantage of the Oracle platform.

Video Highlights:
00:48 Topics for today
01:10 Upcoming Sessions
02:40 Upcoming Sessions early 2022
03:58 Regression Metrics: Agenda Outline
04:17 Regression Metrics: Motivation
04:51 Plot: Target vs Prediction
07:06 Plot: Residual plot
10:20 Mean Squared Error and Root Mean Squared Error
11:45 Mean Absolute Error
12:05 Experiment: MSE/RMSE vs MAE
13:03 Discussion: Weight on Large Error
16:17 R-Squared and Adjusted R-Squared
19:45 Demo on OML Notebook
25:13 Q&A