←
1
2
3
Feature Explorer
Choose the best features for the job.
🎯
Step 1 of 3
What you'll learn in this level
- Selecting features instead of only weighting them
- The difference between classification and regression
- What accuracy, precision and R² say about a model
Feature Explorer
You've learned about weighting features. Now, let's take it a step further: Selecting the right features!
Sometimes, less is more. Adding too many irrelevant features can confuse the model. Your goal is to pick only the most useful features.
How to find good features
- 1
Select a Mission
Pick a dataset: weather, houses or health.
- 2
Pick Features
Pick the features you think matter. Leave out the noise.
- 3
Train & Test
Train your model. Then see how well it works.
- 4
Complete Tasks
Score high on at least 2 datasets. Then you move on.
Think like a data pro. Good features beat many features.
Guess first, then experiment
Two models get the same number of features. Does that make them equally good?