Context:
Employee promotion is key for motivation. JMD Company’s HR wants to use last year’s data to predict promotion eligibility.

Objective:
Predict promotion eligibility.
Identify key promotion factors.
Provide recommendations.

Data:
employee_id: Unique ID
department: Employee’s department
region: Employment region
education: Education level
gender: Employee’s gender
recruitment_channel: Recruitment channel
no_of_trainings: Number of trainings
age: Employee’s age
previous_year_rating: Last year’s rating
length_of_service: Years of service
awards_won: Awards won (1 if yes)
avg_training_score: Average training score
is_promoted: Promotion status (target)

Techniques:
Preprocessing: Handle missing values, encode categories, scale features
Models: Logistic Regression, Decision Tree, Random Forest, Bagging, AdaBoost, Gradient Boosting, XGBoost

Evaluation: Accuracy, Recall, Precision, F1 Score
Tuning: RandomizedSearchCV
Imbalance Handling: SMOTE, RandomUnderSampler

Process:
Preprocess Data: Impute missing values, encode, and scale.
Initial Modeling: Train and evaluate models.
Class Imbalance: Use SMOTE and undersampling.
Tuning: Optimize with RandomizedSearchCV.
Final Model: Select best model based on validation.

Key Findings:
Top Features: Avg training score, age, length of service, previous year rating, awards won.
Department: Higher promotions in Technology, Procurement, Analytics.
Education: Master’s degree holders promoted more.
Region: Some regions have higher promotions.
Gender: No significant difference in promotion rates.

Recommendations:
Training Programs: Improve scores.
Performance Reviews: Ensure consistency.
Awards: Incentivize participation.
Career Planning: Develop clear paths and mentorship.
Standardization: Uniform criteria across regions and departments.
Monitoring: Update model with new data.

Conclusion:
Using this model will streamline promotions, identify high-potential employees, and support fair and data-driven decisions for career development.