This repository contains HR analyst to see Employee resign using Logistic Regression, to help determine the optimum optimum model to see dichotomus value [0,1] as main algorithm. The Python Script is Focused in main thing such : Data Understanding, Data Adjusment, Data insight via Visualization, Modelling, and Deployement Model.
The repository is organized as follows:
data: This directory contains the dataset used for analysis. The dataset includes information about Employee work detail and Factors, like a average hour work, Satisfaction, Salary, Department, etc.
scripts: This directory contains the Python scripts used for data Adjusment, Visualization, Modeling, and Evaluating.
notebooks: This directory contains Jupyter notebooks that provide step-by-step walkthroughs of the analysis process. These notebooks are named and organized by the specific analysis they cover.
The data visualization part of the analysis aims to provide insights into Employee Left Analyzing for Specific Criteria. Visualization techniques such as Kde Plot, bar charts, scatter plot, and Violin are used to see tennure Distribution, Average monthly hours, and another Numerical Distribution. To explore the data visualization analysis, refer to the notebooks in the "notebooks".
Using GridSearchCV to find the best parameters for logistic regression is a common approach to optimize the model's performance. This process will help me find the best hyperparameters for my logistic regression model and improve its performance for my specific dataset and problem.
By examining relevant variables such as job satisfaction, salary, tenure, performance evaluations, and other relevant factors, you can build a predictive model to estimate the likelihood of employees leaving the company.
To evaluate the performance of my logistic regression model using a classification report and accuracy score. These metrics help us understand how well the model predicts employee resign compared to actual values.