The company specializes in executing substantial data science and big data projects as part of their recruitment strategy, offering training courses to potential employees. The primary objective is to determine the genuine commitment of candidates towards long-term employment with the organization after their training. This determination holds significant strategic value, impacting cost reduction, time efficiency, the quality of training programs, and candidate categorization.
The graph illustrates the proportion of employees contemplating a job change versus those who are not. After a thorough analysis of the job-changing issue, the most practical solution is to develop a method for identifying or predicting the likelihood of employees changing their data science job.
As students in group G09, we have undertaken this challenge with a focus on Human Resources (HR) Team involvement, as they are the primary stakeholders significantly impacted by this issue.
Key Information:
- Problem: Identifying employees planning to change jobs post-training.
- Client: Human Resources team of the company.
- Solution: Predicting whether an employee will change their data science job after completing training.
- Goal: Developing a predictive model to assist the HR team in anticipating whether a trainee is likely to seek alternative employment post-training.
We have selected the following public dataset for our project:
This project is structured into 5 main layers:
-
User Interface Layer: This layer focuses on user-friendliness, allowing users to interact with the system. The goal is to implement a simple questionnaire.
-
Data Wrangling and Data Cleansing Layer: This layer is responsible for data cleaning and preprocessing, enhancing the accuracy of the model.
-
Data Mining Layer: This layer analyzes and gathers data using algorithms to transform it into a structured format suitable for further analysis.
-
Model Building and Analysis Layer: Here, predictive models are developed and evaluated to address the business problem.
-
Data Visualizing Layer: This layer provides graphical representations of the dataset to enhance comprehension of model-generated predictions.
- Find a real-world problem and define a solution.
- Data preparation, model building, and training.
- Evaluate the model.
- Make predictions.
- Front-end development and deployment.
Our project begins with dataset selection and data cleaning. We plan to construct two distinct models using different techniques for binary classification. After evaluating and comparing these models, we will develop a user interface (UI) for user interaction.
- Remove columns with less predictive power.
- Handle null values.
- Discretize continuous columns.
- Normalize, reduce dimensions, and integrate data.
- Split the dataset into training and testing sets.
Two models will be developed using the Decision Tree algorithm in Python.
The model will be evaluated using a test dataset, and the best-performing model will be selected based on accuracy.
- Frontend: React.js
- Backend: Streamlit
The primary goal is to provide insights to the HR team regarding data scientists' job changes after training. This assists in informed decision-making, cost reduction, time efficiency, and improved training quality.
- Data quality is assumed to be accurate, complete, and representative.
- Data points are independent.
- The dataset may contain noisy data.
- There are no hidden variables impacting the result.
- The selected features are relevant to the problem.
| Member IT Number | Member Name | Member Role | Member Responsibility |
|---|---|---|---|
| IT21313370 | Dissanayake D.J.R | Team Leader | - Solution Developer |
| - Business Analyst | |||
| - Solution Tester | |||
| - Implement model | |||
| - Handle documentation | |||
| - Test alternate model | |||
| - Data analysis and process | |||
| - UI development | |||
| IT21224348 | Sulakkana H.D.S.R | Solution Developer | - Business Analyst |
| - Solution Tester | |||
| - Implement model | |||
| - Handle documentation | |||
| - Test alternate model | |||
| - Data analysis and process | |||
| - UI development | |||
| IT21302244 | Kuhananth C | Solution Developer | - Business Analyst |
| - Solution Tester | |||
| - Implement model | |||
| - Handle documentation | |||
| - Test alternate model | |||
| - Data analysis and process | |||
| - UI development | |||
| IT21224652 | Manathunga M.A.O.S | Solution Developer | - Business Analyst |
| - Solution Tester | |||
| - Implement model | |||
| - Handle documentation | |||
| - Test alternate model | |||
| - Data analysis and process | |||
| - UI development |