daniellopez46

86 reputation

bio	website
	location	California
	age	33
visits	member for	11 months
	seen	Apr 30 at 15:58
stats	profile views	14

I have several years experience in business and analysis (e.g. descriptive methods using Excel, Access, SQL, and BI tools). I am currently acquiring data mining skills for predictive modeling, specifically Attrition modeling (e.g. finding predictive patterns of turnover for various of our populations, accurately estimating number of terminations by organization and term type).

	86 reputation	bio	website		visits	member for	11 months
17 badges		location	California		seen	Apr 30 at 15:58

summary answers questions tags badges favorites bounties reputation activity

39 Actions

suggestions reviews revisions comments badges posts accepts all

Apr 30	awarded	Notable Question
Jan 14	awarded	Popular Question
Sep 26	comment	Do correlated and/or derived fields require special consideration when using Random Forest? But now I am thinking about our binary target and variables that for the minority class vary but for the majority class are all the same. EX: we are predicting if an employee will continue employment with our company or terminate employment. We have a variable called reason that gives reason why an employee left. For the minority class (i.e "leavers") this varies (i.e. "Better career opportunity", "pay not enough", etc). For the majority class (i.e. "stayers") its all labeled the same thing: "stayed". What about these type of fields?
Sep 26	comment	Do correlated and/or derived fields require special consideration when using Random Forest? That's what I thought but we were not getting an accurate predication with all the variables we had included in the RF model in R. So we took several variables out. Initially we thought it was because of variables like the ones in my original question.
Sep 24	asked	Do correlated and/or derived fields require special consideration when using Random Forest?
Aug 23	asked	What is/are good economic indicator(s) to use for predicting whether or not someone will leave a company?
Aug 3	accepted	How is it possible to turn out with a highly accurate prediction when all records were classified the same way?
Aug 1	revised	How is it possible to turn out with a highly accurate prediction when all records were classified the same way? added screenshots of how we are calc'ing accuracy and also a sample of the data.
Aug 1	comment	Deployment process for Classification models (i.e. decision trees) What is most important to us is accurately identifying and classifying records as "departers" rather than ""stayers" but everyone is classified as "stayers" including the actual "departers" in the data.
Aug 1	awarded	Commentator
Aug 1	comment	How is it possible to turn out with a highly accurate prediction when all records were classified the same way? BTW predicted terms is calculating by summing the propensity scores divided by two. We divide by two since terminations being used to train the model cover a two year period but we are only interested in projecting a year out.
Aug 1	comment	How is it possible to turn out with a highly accurate prediction when all records were classified the same way? Yes, it's a binary classifier and every record is classified as 0 which we call a "stayer" (someone that stayed with the company). I see what you are saying, 2827/3030 is 93% which is close. But the 95%, for example, is calculated actual terminations divided by predicted terms in this case 40/42. Maybe there is some basic logic there or something more I am missing.
Aug 1	revised	How is it possible to turn out with a highly accurate prediction when all records were classified the same way? added confusion matrix
Aug 1	awarded	Quorum
Aug 1	revised	Deployment process for Classification models (i.e. decision trees) added 143 characters in body
Aug 1	asked	How is it possible to turn out with a highly accurate prediction when all records were classified the same way?
Aug 1	asked	Deployment process for Classification models (i.e. decision trees)
Jul 30	accepted	What do Lift and Gain Charts state in the context of an employee turnover model
Jul 30	comment	What do Lift and Gain Charts state in the context of an employee turnover model In other words if the performance is good at each or most deciles compared to baseline then we can expect the total estimate of potential departers to have a certain level of accuracy. Do I more or less have this correct?
Jul 30	comment	What do Lift and Gain Charts state in the context of an employee turnover model @steffen. Thanks for the thorough answer. It is much appreciated. I plan to accept it. What we currently do is take the model and apply it to current population to come up with propensity scores for each individual. We then sum them up the propensity scores by organization to come up with total potential departers. So even in this case were we are not explicitly taking action #2, as you stated, we can still find utility in looking at the gains and lift charts because it does also translate to giving us confidence in our total estimate.