IBM HR Analytics Employee Attrition & Performance
Attrition is a very important metric for businesses to monitor. Retraining is very expensively and timely and replacing people can be very difficult particularly if they had a lot of business process knowledge specific to the company.
If an HR department could predict attrition, they could take steps to that person to reduce the likelihood of them leaving or identify the potential causes that might make someone want to leave the company. Even before hiring an employee, if HR can obtain the prediction result that whether this candidate will leave the company soon or will stay for a long time, it will be an efficient method to control the cost and reduce the risk.
Here we have built a model using the random forest algorithm to predict if someone will leave the company. This methodology and analysis could be of great use to an HR department and is another example of why these methodologies and techniques can be revolutionary in business.
In this report, we used data from IBM HR Analytics Employee Attrition & Performance, which includes a total of 35 variables to apply a machine learning method. The goal of this task is to predict Attrition by selecting a set of explanatory variables and building a random forest classification tree (Supervised Learning).