Job Placement Analysis Project

Table of Contents

• Introduction
• Dataset
• Installation
• Usage
• Modeling Approach
• Results
• Conclusion

Introduction

This project focuses on job placement analysis using data analytics and predictive modeling to optimize hiring decisions based on factors such as academic performance (GPA), work experience, demographics, and salary expectations. The main objectives include:

1. Identifying the most influential factors affecting job placement status.
2. Exploring correlations between demographic attributes (e.g., gender, educational background) and placement outcomes.
3. Analyzing the impact of past work experience and salary expectations on job placements.
4. Providing actionable insights for stakeholders to improve job placement strategies and increase employment opportunities.

This analysis offers personalized career guidance for individuals and data-driven recruitment strategies for organizations.

Dataset

The dataset used in this project contains the following attributes:

• gender: Gender of the candidate.
• age: Age of the candidate.
• gpa: Grade point average (GPA) of the candidate.
• years_of_experience: Years of work experience.
• salary: Salary offered (for placed candidates).
• placement_status: Whether the candidate was placed (1) or not (0).

Data Preprocessing

• Missing values for salary and years_of_experience were replaced with 0.
• Categorical features like gender and placement_status were label encoded.
• Features were normalized using MinMaxScaler.
• SMOTE was used to handle class imbalance in the target variable (placement status).

Installation

1. Clone the repository:

   git clone https://github.com/your-username/job-placement-analysis.git

2. Install the necessary dependencies:

   pip install -r requirements.txt

3. Ensure the dataset job_placement.csv is in the appropriate directory, or modify the path in the script accordingly.

Usage

1. Load the dataset and preprocess the data:

   df = pd.read_csv("job_placement.csv")
   # Preprocessing steps...

2. To visualize the correlation between features:

   corr_matrix = df.corr()
   sns.heatmap(corr_matrix, annot=True, cmap='Reds')
   plt.show()

3. Train and evaluate placement prediction models:

   xc_train, xc_test, yc_train, yc_test = train_test_split(xc, yc, random_state=42, test_size=0.2)
   # Apply SMOTE and train models...

4. Train and evaluate salary prediction models:

   training_regressor(LinearRegression())

Modeling Approach

Placement Prediction

We used the following features to predict whether a candidate would be placed:

• gender
• age
• gpa
• years_of_experience

After data preprocessing, SMOTE was applied to handle imbalanced classes. The features were normalized, and models such as Random Forest, Extra Trees, Gradient Boosting, AdaBoost, Linear Regression, and XGBoost were evaluated.

Salary Prediction

For candidates who were placed, salary predictions were made using the same models listed above. The salary values were predicted based on:

• gender
• age
• gpa
• years_of_experience

The models were evaluated using Mean Absolute Error (MAE) and Mean Squared Error (MSE) metrics.

Results

Model	MAE	MSE
Random Forest	0.82	1.14e-07
Extra Trees	6.18e-12	7.35e-31
Gradient Boosting	1.49	1.16e-07
AdaBoost	240.47	0.0013
Linear Regression	1.59e-12	2.97e-31
XGBoost	0.81	1.47e-08

Key insights:

• GPA and years of experience were significant predictors of job placement.
• Gender and salary expectations showed varying impacts on placement outcomes.
• The Extra Trees Regressor and Linear Regression models performed best for salary prediction, with minimal error.

Conclusion

This project highlights how data analytics and machine learning can be applied to job placement analysis. By identifying the key factors influencing placement success, this project can help individuals receive personalized career advice and enable organizations to make informed recruitment decisions, fostering a skilled and diverse workforce.