Designed a web platform enabling users to interactively manipulate data generation processes and model hyperparameters, with visual feedback of the outcomes
Demonstrated knowledge and proficiency in the inner workings of various classification algorithms
Utilized Matplotlib for plotting the data and visualizing the decision boundaries
Developed an API using Flask to facilitate plot generation
Developed a spotify scraper and a corresponding website enabling users to receive personalized recommendations based on Spotify track links or IDs
Utilized a custom optimized K-Nearest Neighbors algorithm for recommendation generation
Implemented MongoDB as the database backend for efficient data storage and flexible querying capabilities
Containerized the web server using Docker for seamless deployment and execution across diverse environments
Leveraged the Spotify API to scrape the audio features of over 7 million songs
Implemented robust error handling within the scraper as to prevent interruptions due to errors during operation, thereby ensuring continuous operation over extended durations without interruptions
Demonstrated adeptness in utilizing both, official and unofficial APIs for data acquisition
Designed a minimalist yet responsive frontend interface, prioritizing usability and simplicity without unnecessary clutter