Prediction Playground
Fit a straight line through two columns, then predict a new value.
Demonstration using a representative 20-row sample dataset: a simple linear fit, not a production ML model.
Predicted Y: –
What is happening?
This fits the simplest possible model (a straight line, y = slope×x + intercept) using ordinary least squares. R² (0 to 1) says how much of Y's variation the line actually explains; a low R² means the line is a weak predictor even if it's the best straight line available. This is a teaching tool, not a forecasting system: real prediction problems need more than one input and a lot more validation.
Under the Hood
X / Y column selection
↓
Ordinary least squares fit
↓
Line drawn over the scatter, R² reported
↓
Custom X evaluated through the fitted line
Inputs
- 5 numeric columns, 20 rows
Processing
- Least-squares linear regression
- R² goodness of fit
Output
- Fitted line + equation
- Live prediction for a custom X
Technology
- Vanilla JavaScript
- Hand-built inline SVG chart
Explore another experiment
More Playground experiments are on the way.