1. Blind - for - Students | GitHub
Tech Stack: Next.JS, PostgreSQL, Prisma, Typescript
○ A blind like application that allows students to post doubts or queries anonymously.
○ PostgreSQL as a database along with Prisma to simplify CRUD operations, Next.JS as a full-stack react framework and TailwindCSS for styling components.
2. Dog - Breed - Identification | GitHub
Tech Stack: Tensorflow, Google Colab
○ Utilized Tensorflow and Keras to develop a convolutional neural network (CNN) for identifying 120 different dog breeds from images.
○ Achieved a top accuracy score of 0.65126 on the competition leader-board. Employed transfer learning with pre - trained models to improve prediction accuracy and model robustness.
3. Bulldozers - Price - Prediction | GitHub
Tech Stack: Scikit - Learn, Jupyter Notebook
○ A machine learning model to analyze and predict the prices of bulldozers based on historical data.
○ Analyzed 52 variables and handled missing data with various methods and used over 400,000 rows of data for training and validation.