Challenge_Round_8

HR Analytics: Predict Employee Attrition

Challenge description:

You work as a data scientist for a pharmaceutical company, and in the last two years several employees have left the organization for unpredictable or uncontrollable reasons. The Head of HR is very worried because frequent recruiting processes are slow, costly and hinder the growth of the company. Hence, she asks you to investigate the phenomenon, and come up with good recommendations to reduce employees' attrition rate.

First, she wants to receive an interactive and customizable dashboard that provides a highly informative overview of the employees in the company. Mind that she does not want to be overwhelmed with too many numbers or visualizations, so you should pick wisely what you present.

Second, you should build a machine learning pipeline to predict employees' attrition rate. Take the preprocessing steps that you deem necessary and adopt a machine learning algorithm of your choice. Clearly, she would like to obtain predictions that are as accurate as possible. Additionally, she expects you to be able to explain the model's decision-making process (identifying crucial factors that lead employees' to leave their job), and provide recommendations to increase employees' happiness and retention rate.

Key requirement: you must use an explainable AI (XAI) technique of your choice to explain the model's predictions and provide a short written description (max. 100 words) in an annotation. For example, you could use one of KNIME Verified Components on Model Interpretability: https://hub.knime.com/knime/spaces/Examples/00_Components/Model%20Interpretability~WMtQn1U91a-xzZY3/.

Outcome: 

An interactive dashboard to explore employees' key characteristics and explain employees' attrition using an explainable AI (XAI) techniques.

Deliver your solution as a separate workflow and name it: Solution_Round_8_<your_team_name>. Place your solution workflow in the same folder of this challenge workflow.

Teams are strongly encouraged to submit high-quality work in order to improve their chances of getting maximum points. Don't be afraid to go the extra mile! :)

Dataset:

IBM HR Analytics Employee Attrition dataset from Kaggle: https://www.kaggle.com/datasets/pavansubhasht/ibm-hr-analytics-attrition-dataset?resource=download

Deadline:

March 24, 2024 (submission by 11:59 PM CET) **. Check the calendar of the tournament: https://info.knime.com/game-of-nodes

** We will verify the date and time of the latest edits.

KNIME Game of Nodes:

Rules, Assessment Criteria & FAQs: https://info.knime.com/game-of-nodes

External resources

• KNIME Game of Nodes
• IBM HR Analytics Employee Attrition dataset