In this article we will get an understanding of Sum of squared residuals
When learning backpropagation, this is a rule that we have to use often because it defines how we measure and reduce error in a model.
What is a residual?
Suppose you have some real data points and a line that tries to fit them.
- Actual value → what really happened
- Predicted value → what your model says
The residual is:
residual = actual - predicted
So, it tells you how wrong your prediction is for one data point.
Reason for squaring the residuals
This makes all the errors positive and also penalizes large errors more than small ones.
Sum of squared residuals (SSR)
Now, when you do this for every data point and add them together, you will get this equation:
SSR = ∑(actual − predicted)²
This is called the sum of squared residuals, which measures the total error of the model.
The rule
Here, the rule is basically this:
The best-fitting model is the one that minimizes the sum of squared residuals.
This is known as the Least Squares Principle.
Wrapping up
This is another byte-sized piece to know before going into backpropagation.
If you’ve ever struggled with repetitive tasks, obscure commands, or debugging headaches, this platform is here to make your life easier. It’s free, open-source, and built with developers in mind.
👉 Explore the tools: FreeDevTools
👉 Star the repo: freedevtools
Top comments (0)