I am a Master’s student at the University of North Carolina in Charlotte majoring in Computer Science. Superior understanding of reporting tools, including scorecards, graphs and charts. Proficient in Python, SQL and Tableau. Comfortable collaborating with team members and working independently. Talent for conveying a clear and compelling narrative and enthusiasm for creating vital functionality. Strong desire for continuous improvement and proven ability to think strategically and act tactically about product development.
-
Experience
Worked as a Statistics Teaching Assistant where I tutored students with statistics concepts, assist the professor with the coursework, oversee assignments and grade them.
-
Projects
Auto Insurance Loss Ratio Prediction (Python, Jupyter Notebook, Orange) August 2019 – December 2019
• Participated in a Kaggle Competition under the Big Data Analytics course where the objective is to predict the loss ratio with high accuracy results having more than 420000 training data
• Predicted the target variable in a supervised learning problem which is the natural logarithm of the loss ratio of portfolios
• Conducted iterative data exploration and transformation to make data credible and useful in Orange and Python
• Initiated with a simple Linear Regression, moving forward with other complex models such as Decision Tree Regressor, Random Forest Regressor, Support Vector Machines, and LightGBM
• Tested the model using the holdout technique and the cross-validation and tried to minimize the Mean Absolute Error (MAE), Root Mean Square Error (RMSE) and the R Squared error for which LightGBM turned out to be the best
Food Delivery System (MySQL, MySQL Workbench, draw.io) August 2019 – December 2019
• Led a team to design a fully normalized database system using business rules, entity-relationship diagram, normalization, and schema design using MySQL and MySQL workbench
• Integrated Triggers, Stored Procedures and Views which employ joins, nested selects, conditions, and sorting
Music Genre Classification (Python, Jupyter Notebook) January 2019 – May 2019
• Implemented a Machine Learning project on a GTZAN dataset having 10 genres each containing 100 audio files
• Pre-processed the data which includes converting it to melspectrograms and generated charts and reports for analysis
• Built models such as, KNN, Support Vector Machine, Convolution Neural Networks, Auto-encoders from the scratch to analyze the number of layers and different activation functions that work well for better accuracy
• Presented a poster and explained the project to a non-technical crowd
Blackjack Player (Python, Jupyter Notebook) March 2019 – April 2019
• Developed a Reinforcement learning agent to solve the blackjack problem
• Explored Temporal Difference(TD) Learning methods and executed SARSA (state-action-reward-state-action)
• Optimized the game and the best possible winner is chosen by updating the Q-tables. The rewards obtained are learned and suitable action is chosen to generate good outcomes by the agent
Automated Generic Medicine Prescription (Node.js, Php, Javascript, HTML, CSS, SQL) August 2017 – May 2018
• Deployed a web application using Node.js that gives a cheaper generic alternative to the branded medicines prescribed by the doctors; by checking substituents such as active ingredients of prescribed medicine or health symptoms
• Engineered an algorithm (Trie data structure) that improved data retrieval rate by 99.30% compared to the traditional data retrieval