DAndresSanchez/Placements_Multiple_Linear_Regression

Work placement salaries analysis through multiple linear regression and their occurrence based on qualifications and work experience.

Work Placements analysis through Multiple Linear Regression

Data source

https://www.kaggle.com/benroshan/factors-affecting-campus-placement

Description

Salary prediction based on grades and education

• File: salary_linear_regression.py
• Analysis of the data to predict salary from education.
• Use of Multiple Linear Regression.
• Comparison with Ridge and Lasso.

Work placements prediction based on education

• File: placement_classifier.py
• Analysis of the data to predict the occurrence of a work placement.
• Use and comparison of Random Forest Classifier, Logistic Regression and KNeighbors Classifier.

Revised skills

• Data import from .csv files
• Data cleaning with pandas
• Visualisation in Matplotlib
• Supervised learning: classifiers with scikit-learn

Results

Salary prediction based on grades and education

• Salary prediction vs. Actual salary

• Feature selection with Lasso

Work placements prediction based on education

ROC curves for:

• Logistic Regression

• K-Nearest Neighbors

• Random Forests