Implement the ASD metric from the book René Michel, Igor Schnakenburg, Tobias von Martens. Targeting Uplift. An Introduction to Net Scores.
Where: metrics module
Explanation
📌 The ASD (average squared deviation) is a model stability metric shows that model does not overfit on training data (extremely important in uplift modeling).
Explanation of what are l^T and l^V you can get from the table:
(page 103)
This table 4.1 is a source for the ASD metric. Table is already implemented as metrics.uplift_by_percentile function.
📌 Realization of the metric comes down to the next steps:
call uplift_by_percentile twice: for train data and for the validation data
Mogby
commented
3 years ago
💡 Feature request
Implement the ASD metric from the book
René Michel, Igor Schnakenburg, Tobias von Martens. Targeting Uplift. An Introduction to Net Scores.Where:
metricsmoduleExplanation
📌 The ASD (average squared deviation) is a model stability metric shows that model does not overfit on training data (extremely important in uplift modeling).
Explanation of what are l^T and l^V you can get from the table:
metrics.uplift_by_percentilefunction.📌 Realization of the metric comes down to the next steps:
uplift_by_percentiletwice: for train data and for the validation dataAdditional context
Please provide screenshots with working examples.