Cost-sensitive error rate

The problem, from first principles

Error rate treats every mistake as equal. In practice a missed disease costs far more than a false alarm. A cost matrix states what each kind of mistake costs, and the cost-sensitive error is the average cost per sample.

From theory to code

Implement cost_sensitive_error(y_true, y_pred, cost_matrix), returning the mean cost over all samples.

Constraints

• y_true and y_pred are integer class arrays of equal length.
• cost_matrix[i][j] is the cost of predicting j when the truth is i; correct predictions normally cost 0.
• Return a Python float; return 0.0 for empty input.

Hints

Open one at a time. Each gives away a little more than the last.