Tunisian Fraud Detection Challenge
Helping Tunisia
Knowledge
Active
Classification
Anomaly Detection
1443 joined
276 active
Tip to get the real tax adjustment
Notebooks · 20 Sep 2019, 17:39 · 2
we had applied the log(1 + x) for the target variable. In order to get the real value, you can apply its inverse function which is the exp(x) - 1.
thank you for this, I am working on the data offline, is this transformation the result of research or did the data owners emphasize this in the competition in place.??
Hi Bennour,
Log transformation are convenient means of transforming a high skewed variable into a more normalised one.
(x) is equivalent to 10% increase in Xi
Variables with non-linear relationships, the chances of producing errors may also be skewed negatively. It is better to reduce the skewness.