I am reaching out to seek guidance on forecasting employee attrition using machine learning techniques. While I've come across numerous resources, I believe insights from the community could significantly enhance my approach.
Here's a brief overview of what I am trying to do:
Data Availability: I have 10 years' worth of historical data encompassing various demographics such as age, years of service, education, work location, department name, job title, etc. Additionally, the dataset includes termination dates and attrition indicators.
Data Organization: The data is structured within an Excel file; all historical data are gathered in one Excel worksheet, with the variable "Period’ containing years. The same employee ID would be multiple times per year if the employee is still working.
Given this context, I have several questions:
Data Transformation: What methodologies should I employ to transform the data into a format suitable for machine learning models? Especially with the historical data and the fact the employee ID can be on the file multiple times. Does this affect modelling?
Result Exportation: Once the model generates predictions, what's the recommended approach for exporting the file containing the results?
Future Forecasting: How can I effectively forecast employee attrition for the next 10 years based on historical data, leveraging machine learning methodologies?
Your expertise and insights on these matters would be immensely valuable as I navigate this endeavour. Any advice, best practices, or suggested resources you could offer would be sincerely appreciated.
