Hi there! 👋 In this repository, I summarise some of the projects I have worked on, including data analysis, visualisation, machine learning, and the business insights I have gained.
- Gmail: febbyyanggraini01@gmail.com
- Medium: medium.com/@febbyngrni
- LinkedIn: linkedin.com/febbyanggrainii
Credit Risk Prediction - Report
This project aims to predict potential loan defaults to help Bank Republik reduce Non-Performing Loans (NPL). By optimizing the credit scoring system, the model prevents high-risk borrowers from being approved. The model achieves 87% accuracy, with an improved recall for credit default cases from 76% to 81%, ensuring that more high-risk applicants are correctly identified.
Country Clustering For Effective Aid Allocation - Report
This project helps HELP International, a NGO, optimize the allocation of $10 million in aid by identifying countries with the highest need. Using K-Means and DBSCAN clustering, 167 countries were grouped based on socio-economic indicators to prioritize funding distribution. The analysis identified 3 optimal clusters, with 48 countries classified as high-priority for aid.
Coffee Shop Executive Dashboard
This project provides real-time revenue tracking for PT Kopi Kenalan, helping executives monitor sales performance. The dashboard highlights underperforming stores and area managers to ensure revenue targets are met. The forecast indicates that the company is on track to meet its target in 22 days, but a decline to the lower bound of historical trends may require strategic intervention.
Telco Churn Prediction - Report
This project helps IndiHouse, a telecom company, predict customer churn and implement targeted retention strategies. By identifying high-risk customers early, the company can take preventive actions. The model achieved 74% accuracy, with a churn recall of 72%, ensuring better detection of at-risk customers. The solution is deployed using AWS EC2 for real-time churn prediction.
Flight Fare Prediction - Report
This project helps EaseMyTrip, an OTA, optimize ticket pricing using machine learning to maximize revenue and stay competitive in a fluctuating market. By predicting fares based on factors, the company can implement dynamic pricing strategies effectively. The Random Forest model achieved the best performance with 98% R² and RMSE of 2668.17, accurately capturing price variations.
RFM Analysis: Customer Segmentation - Report
This project helps a retail company stabilize revenue by identifying customer segments through RFM analysis. By understanding purchasing behavior, the company can develop targeted marketing strategies to boost retention and sales. K-Means clustering revealed three segments. These insights enable personalized promotions to enhance engagement and revenue.
Marketing A/B Testing - Report
This project evaluates the effectiveness of an ad campaign by measuring its impact on conversion rates through A/B Testing. By comparing ad exposure to a PSA, the company aims to refine its marketing strategy for better ROI. Statistical analysis (Z-test: 6.1291, p-value: 0.0000) confirms that the ad significantly increased conversion rates.
- Designing Relational Database: Used Cars Database System - Report
- Dashboard: Superstore Sales Analysis - Report
- Dashboard: HR Attrition Analysis - Report
- Data Exploration: Crime in Los Angeles - Report
-
Data Collection and Storage: MySQL and PostgreSQL, Spreadsheet .
-
Data Processing and Analytics: Jupyter Notebook, Pandas, Numpy.
-
Development: Python and Git.
-
BI Tools: PowerBI, Tableau, Looker Studio.
-
Data Visualization: Matplotlib and Seaborn.
-
Machine Learning Modeling: Classification, Regression, Clustering.
-
Machine Learning Deployment: FastAPI and Docker.