Comments on: A Tutorial on People Analytics Using R – Employee Churn

By: Laura Valenzuela

Laura Valenzuela — Mon, 07 Aug 2017 19:30:51 +0000

Hi, Lyndon:

Thank you for the post.

I am very confused. Why the error matrices of Decision Tree, Boosted Model and SVM model are EXACTLY the same? That does not sound logical to me. Please, please, help me to understand.

Laura Valenzuela

By: Lanny Reay

Lanny Reay — Wed, 03 May 2017 09:23:58 +0000

Exactly what I was looking for, thank you for posting.

By: Mandyc

Mandyc — Mon, 27 Mar 2017 23:49:37 +0000

Excellent post. I was checking constantly this blog and
I am impressed! Extremely useful information particularly the closing phase 🙂 I handle such information a lot.
I was seeking this certain info for a long time. Thank you and best
of luck.

By: Kumar

Kumar — Fri, 13 Jan 2017 12:07:07 +0000

Great Work Sir!! Thanks

It will be a helpful to understand step by step approach in analysing the data for New R programmer like me .

Thanks again…Kudos to you.!!

By: Neeraj Pandey

Neeraj Pandey — Wed, 11 Jan 2017 05:47:41 +0000

Excellent tutorial !!!
I am also working on the Employee attrition model and has given me good points to cover.
May I know have you also worked on the Employee Profiling and predict no of day for the employee attrition, if yes please share that too

By: Lyndon Sundmark

Lyndon Sundmark — Wed, 07 Sep 2016 02:27:14 +0000

Hi Bennet. This was my first attempt to bring predictive analysis down to individual level ,rather than just leaving it a turnover ‘rate’ analysis in the aggregate, so i cant speak from the individual data level. But in terms of sensitivity and predictive value at the turnover rates level of analysis, I have found typically that turnover is sentitive to age, years of service sometimes , and to other organizational descriptive information- such as dept, division, and sometimes ‘reports to’. Every organization will be different. My purpose for sharing the example to encourage others to apply a similar data science approach on their organizations data, to have their data speak to them. Very few generalizations I can make beyond that, because organizations can be different. Your observation below could very much be the case. Having said that, it may too be the human behaviour or others that leads to some peoples termination decisions. Hope that helps, and thank you for taking the time to share your thoughts. It is appreciated

By: Bennet

Bennet — Tue, 06 Sep 2016 15:56:02 +0000

Hi,

This is a fantastic tutorial.

You’ve used synthetic data for the demonstration; in the real world, what have you seen in terms of “great” model sensitivity and predictive value?

Obviously one could add many more features to the list of independent variables to increase predictive power, but at the end of the day we are predicting human behavior–I suspect even the best models have some kind of type II error (which would be the worst kind either from a talent management or workforce planning lens).

Cheers,

Bennet