Overview
Regression metrics evaluate the performance of regression models by comparing predicted values to true values.
Functions
mse
Calculate Mean Squared Error (MSE).
function mse(yTrue: Tensor, yPred: Tensor): number
Ground truth (correct) target values
MSE value (always non-negative, 0 is perfect)
import { mse, tensor } from 'deepbox/metrics';
const yTrue = tensor([3, -0.5, 2, 7]);
const yPred = tensor([2.5, 0.0, 2, 8]);
const error = mse(yTrue, yPred); // 0.375
rmse
Calculate Root Mean Squared Error (RMSE).
function rmse(yTrue: Tensor, yPred: Tensor): number
Ground truth (correct) target values
RMSE value (always non-negative, 0 is perfect)
import { rmse, tensor } from 'deepbox/metrics';
const yTrue = tensor([3, -0.5, 2, 7]);
const yPred = tensor([2.5, 0.0, 2, 8]);
const error = rmse(yTrue, yPred); // √0.375 ≈ 0.612
mae
Calculate Mean Absolute Error (MAE).
function mae(yTrue: Tensor, yPred: Tensor): number
Ground truth (correct) target values
MAE value (always non-negative, 0 is perfect)
import { mae, tensor } from 'deepbox/metrics';
const yTrue = tensor([3, -0.5, 2, 7]);
const yPred = tensor([2.5, 0.0, 2, 8]);
const error = mae(yTrue, yPred); // 0.5
r2Score
Calculate R² (coefficient of determination) score.
function r2Score(yTrue: Tensor, yPred: Tensor): number
Ground truth (correct) target values
R² score (1 is perfect, 0 is baseline, negative is worse than baseline)
import { r2Score, tensor } from 'deepbox/metrics';
const yTrue = tensor([3, -0.5, 2, 7]);
const yPred = tensor([2.5, 0.0, 2, 8]);
const score = r2Score(yTrue, yPred); // Close to 1 for good fit
adjustedR2Score
Calculate Adjusted R² score.
function adjustedR2Score(yTrue: Tensor, yPred: Tensor, nFeatures: number): number
Ground truth (correct) target values
Number of features (predictors) used in the model
import { adjustedR2Score, tensor } from 'deepbox/metrics';
const yTrue = tensor([3, -0.5, 2, 7]);
const yPred = tensor([2.5, 0.0, 2, 8]);
const score = adjustedR2Score(yTrue, yPred, 2); // Adjusted for 2 features
mape
Calculate Mean Absolute Percentage Error (MAPE).
function mape(yTrue: Tensor, yPred: Tensor): number
Ground truth (correct) target values
MAPE value as percentage (0 is perfect, lower is better)
import { mape, tensor } from 'deepbox/metrics';
const yTrue = tensor([3, -0.5, 2, 7]);
const yPred = tensor([2.5, 0.0, 2, 8]);
const error = mape(yTrue, yPred); // Percentage error
function medianAbsoluteError(yTrue: Tensor, yPred: Tensor): number
Ground truth (correct) target values
Median absolute error (always non-negative, 0 is perfect)
maxError
Calculate maximum residual error.
function maxError(yTrue: Tensor, yPred: Tensor): number
Ground truth (correct) target values
Maximum absolute error (always non-negative, 0 is perfect)
import { maxError, tensor } from 'deepbox/metrics';
const yTrue = tensor([3, -0.5, 2, 7]);
const yPred = tensor([2.5, 0.0, 2, 8]);
const error = maxError(yTrue, yPred); // 1.0 (worst prediction)
explainedVarianceScore
Calculate explained variance score.
function explainedVarianceScore(yTrue: Tensor, yPred: Tensor): number
Ground truth (correct) target values
Explained variance score (1.0 is perfect, lower is worse)