Process IBM Employee Attrition

rapid1234

Hi, 

I am new and working with the dataset employee attrition. My Accuracy is not the best (84%). I tried many algorithms but that's the best accuracy. The standard deviation is also nod good (1,5).
How can I improve my accuracy and  standard deviation? Are this values good?

Does someone have a finished process with the data set employee attrition for impressions?

Thank you!

Caperez

Hi @rapid1234,

The standard deviation is highly influenced by outliers, so a good practice is to remove the outliers of your dataset. Also a bigger dataset is a good practice to for more realistic standard deviation.

an empirical rule for a normal data distribution is that approximately 68% of the values fall within one standard deviation of the mean,

Best