• Ranked top 50 among 280 teams
• Achieved AUC: 76% (XGBoost), 72% (Random Forest)
• Selected 50 key features via MI Score & Chi-Squared tests
• Used Isolation Forest to remove outliers, enhancing prediction reliability

Skills: Python, NumPy, Pandas, Machine Learning, Predictive Modelling, Data Preprocessing, Feature Importance, XGBoost, Random Forest, Isolation Forest, Model Evaluation

• Investigated survival rates across different passenger classes
• Examined the relationship between age and survival, highlighting survival disparities across age groups
• Analyzed gender disparities, showing a significantly higher survival rate for women compared to men
• Explored the impact of ticket fare and embarkation points on survival rates
• Applied EDA techniques including heatmaps, histograms, boxplots, scatter plots, and pie charts for data visualization
• Used the Interquartile Range (IQR) method to identify and remove outliers
• Preprocessed the data by handling missing values and encoding categorical variables

Skills: Python, Pandas, NumPy, Matplotlib, Seaborn, Data Preprocessing, EDA, Data Visualization

• Visualized Netflix's content distribution across movies and TV shows.
• Created interactive world map to show geographical content presence.
• Analyzed content ratings, with TV-MA and TV-14 being most common.
• Developed genre popularity insights, highlighting top 10 genres.
• Illustrated content growth from 2007 to 2021, showing significant expansion post-2015.

Skills: Tableau, Data Visualization, Data Analysis, Dashboard Design, Data Cleaning, Business Intelligence

• Identified peak revenue hours and seasonal demand to enhance pricing strategies
• Segmented casual vs. registered riders to improve membership growth
• Outlined future enhancements including weather impact analysis

Skills: SQL, Power BI, Data Visualization, Data Analysis

• Processed data using Python (Pandas) for accuracy
• Visualized insights with bar charts and word clouds
• Identified key business opportunities through data analysis

Skills: Python, Pandas, Power BI, Excel, Data Analytics, Data Visualization

• Preprocessed Fashion MNIST dataset, applying normalization, random rotation, and horizontal flipping to enhance model generalizability
• Designed and implemented a CNN architecture in PyTorch, incorporating convolutional layers, batch normalization, ReLU activations, and max pooling
• Achieved a validation accuracy of 91.97% and an F1-score of 0.9192, demonstrating strong classification performance
• Utilized Optuna for hyperparameter tuning, optimizing dropout rates, learning rates, and number of epochs
• Visualized model embeddings using UMAP to analyze class separability and identify misclassification patterns
• Compared CNN performance with CLIP zero-shot classification, highlighting limitations in grayscale image processing

Skills: PyTorch, CNN, Image Classification, Optuna, Data Augmentation, UMAP, CLIP, Hyperparameter Tuning, Deep Learning

• Architected and hosted AWS EC2 Linux instance with MySQL, reducing database setup time by 30% by automating configurations and leveraging command-line tool
• Designed and implemented database schema with foreign key constraints and executed ETL processes, ensuring data integrity and streamlined loading
• Enhanced database availability to 98% by implementing a peer-to-peer replication strategy and configuring security groups on AWS
• Automated eCommerce sales data updates with MariaDB stored procedures and triggers, reducing update time by 98% and improving operational efficiency
• Migrated SQL to NoSQL (MongoDB), optimizing for scalability and flexibility in handling unstructured data
• Analyzed MongoDB JSON documents to extract insights on customer behavior and inventory management, driving strategic business decisions

Skills: SQL, Triggers, Stored Procedures, ETL, AWS EC2, MongoDB, MariaDB, OrientDB, Clustering, Peer-to-Peer Replication, CAP Theorem, Data Management

• Designed and developed the Product Canvas for "DeliverEase," a conceptual app focused on comparing food prices and delivery times across multiple delivery services
• Conducted in-depth research to identify and define key user personas, aligning product features and design with user needs and pain points
• Determined critical product metrics to track performance, including user engagement, conversion rates, and delivery satisfaction
• Created mockups for essential app flows, including onboarding, search, compare, and checkout, ensuring an intuitive user experience

Skills: Figma, draw.io, Product Management, User Personas, User Stories, UX Research, UML, Mockups, Wireframes, Product Roadmap, User Stories