-
Machine Learning-Based Credit Card Fraud Detection with XAI
This project uses advanced machine learning techniques to handle credit card fraud detection, leveraging a dataset of 284,807 transactions. Employing Logistic Regression and XGBoost models, the project achieves notable accuracy and interpretability, with the fine-tuned XGBoost model reaching an accuracy of 99.97% and a ROC AUC score of 98.31%. Key to the project is the application of Synthetic Minority Over-sampling Technique (SMOTE) to address class imbalance and the use of SHapley Additive exPlanations (SHAP) for understanding feature influence, providing a comprehensive approach to identifying fraudulent transactions and offering insights into the mechanics behind the model's decisions.
GitHub Repo | Learn More -
NLP-Driven Fake News Detection
This project delves into the challenge of fake news detection using Natural Language Processing (NLP) techniques. Utilizing Python, the project explores machine learning and deep learning methodologies to distinguish between authentic and fake news. The project underscores the significance of data collection, pre-processing, and feature extraction in the successful application of NLP techniques in machine learning and deep learning.
GitHub Repo | Learn More -
Data-Driven Insights into Global Terrorism: A Quantitative Analysis
This research project leverages the power of Python to dissect the Global Terrorism Database (GTD). Using a quantitative approach, the study uncovers patterns and trends in global terrorism incidents through descriptive statistics. The project deliverables include a comprehensive report that outlines the research design, data collection methods, and analysis techniques. The findings emphasize the critical role of data visualization in making sense of complex datasets.
GitHub Repo | Learn More -
Statistical Analysis and Predictive Modeling of Customer Complaints in Banking
This project demonstrates my ability to conduct a thorough statistical analysis of a banking dataset with the aim of understanding and predicting customer complaints. The analysis identified key customer demographics more likely to file complaints, explored the non-significance of credit scores in predicting complaints, and assessed the bank's reward system. A logistic regression model was developed, achieving an accuracy of 99.95% in predicting customer complaints.
GitHub Repo | Learn More -
Object-Oriented Programming (OOP) Project: Banking System Design with Python
This project is a prototype of a banking system for ABCBank, demonstrating the use of Object-Oriented Programming (OOP) principles and design patterns in Python. The system supports various types of accounts, account upgrades, and loan applications. Design patterns used include Strategy, Abstract Factory, Decorator, and Facade, providing a robust, flexible, and extensible system.
GitHub Repo | Learn More -
Object-Oriented Programming (OOP) Project: Data Imputation Using Strategy and Factory Patterns
This project was part of a school assignment to demonstrate proficiency in Object-Oriented Programming concepts. This project involved creating a custom Imputer class to handle missing data in datasets, using the Strategy and Factory design patterns in Python. The Imputer class supports mean, median, and mode imputation strategies. The Strategy pattern allowed for flexible selection of imputation methods, while the Factory pattern ensured robust and error-free instantiation of strategy objects.
GitHub Repo | Learn More -
CPU Scheduling Algorithms Simulation
This project shows the implementing and analyzing of various CPU scheduling algorithms, including First-Come, First-Served (FCFS), Shortest Job First (SJF), Priority Scheduling, and Round Robin. The project involves the creation of Python classes for each scheduling algorithm, simulating their behavior with a dataset of processes. The analysis provides insights into the performance of each algorithm, such as wait times and execution times.
GitHub Repo | Learn More -
Predictive Modeling and Demand Analysis of NYC Taxi Fares
This study uses historical data to analyze NYC taxi fares, identifying key factors influencing fare amounts and predicting fares using a linear regression model. The analysis reveals weekly demand patterns and peak operation times, providing insights for optimizing taxi services and fare estimation.
GitHub Repo | Learn More -
Optimizing Productivity in Garment Manufacturing
This project presents a comprehensive data analysis of a garment manufacturing company, focusing on productivity levels, employee incentives, and operational efficiency. Utilizing real-world data, the study uncovers key insights into team performance, the impact of rest days, and the effect of style changes on productivity. The findings provide a foundation for developing targeted strategies to enhance overall productivity.
GitHub Repo | Learn More -
Software Development Practices: From Requirements Engineering to Version Control
This term paper offers an in-depth exploration of various aspects of software development, including requirements engineering, system design, software testing methodologies, software development life cycles, version control, and practical usage of Git. The paper provides a detailed case study of a train ticket booking system, demonstrating the application of theoretical concepts in a real-world scenario. It also includes a comparative analysis of waterfall and agile methodologies, emphasizing their respective advantages and disadvantages. This work serves as a valuable resource for both aspiring and experienced data scientists looking to understand and apply best practices in software development.
GitHub Repo | Learn More