to use F1 score: when classes are imbalanced and both FP and FN matter

Why might a sports team need a different score to judge their performance?

Imagine a soccer team where most games are won by a single goal margin, but occasionally they lose by a large margin. The team wins most games but struggles when they lose.

In soccer, if wins and losses are equally important, a simple win-loss record doesn't reflect the team's true performance. We need a score that considers both winning and losing situations, especially when losses are rare but impactful.

Example

A team wins 9 games and loses 1 by 5 goals. A simple win-loss record shows 10 wins, but it doesn't show the importance of losing 5 goals.

Remember this

The F1 score helps balance the importance of wins and losses, especially when losses are rare but significant.

Text adapted from Wikipedia, licensed under CC BY-SA 4.0.

Related concepts

Precision and recall

Precision = Relevant retrieved instances / All retrieved instances

to standardize: when you need zero mean and unit variance for gradient-based optimization

Why do we need to make data uniform before training a model?

to use XGBoost: for tabular data where you want the best possible performance

Why do some people always seem to win at chess?

denoising score matching does: learns to denoise, which equals learning the score

Propensity score matching (PSM) reduces bias in treatment effect estimates

to normalize features: when features have different scales and you use distance-based methods

Why do some things need to be adjusted to compare fairly?

Bias vs variance: high bias = underfitting, high variance = overfitting

Can a perfect fit to past data predict future events?

Swipe through 100 ML concepts daily

Open Pocket Polymath