Click Icon For GitHub Link
Sample Research Project:
https://github.com/miyading/tempodetermination
Sample Course Projects:
https://github.com/miyading/courses
​
A user-centered Repertoire APP with Graphical User Interface. The Repertoire APP is designed to enhance a musician's day-to-day practicing. The musician can load their repertoire, make changes to a specific piece, add more pieces as assigned, listen to a particular piece before practicing, and save their daily progress using this APP. For each piece in the repertoire, the user can specify its anticipated tempo and its estimated practice time, and label its urgency/priority in order to plan out a better practice schedule. The APP is currently capable of playing any of 443 the midi audio files (Bach Chorales, Chopin Preludes, or Joplin Rags) in the repository upon request. Additional features include a visual metronome, as well as an additional frame to retrieve further web analysis for the user's favorite movement of any Mozart Sonatas of the day (KernScores). The construction adheres to software design principles and has been tested for its robustness.
​
Please find below a little demonstration of saving a piece and playing a midi recording (Scott Joplin's Maple Leaf Rag) using my Repertoire APP, as well as other functionalities:
CPSC 210 Personal Project (2019 Winter)
STAT 344 Group Project (2020 Winter)
During the current pandemic, all of us are adjusting to online courses. Therefore, in the group project for STAT 344 Sample Surveys, we applied our knowledge of sampling strategies to infer if the average UBC class sizes are well-situated to be potentially conducted in-person with regards to the Provincial Protocols. Over two months, our group members collaborated online to collect the data, calculate appropriate sample sizes, and implement our two sampling schemes: SRS and Stratified Sampling with Proportional Allocation. During the next phase, we analyzed the dataset and generated plots using R, synthesized different estimations and sampling errors into tables, and each took parts of the writing using Latex. Please find the final report and associated R codes in the STAT 344 Group Project.zip file.
Sample Summer Work Learn Projects:
https://github.com/miyading/final-report
Hackathon Project
Leading teammate on two rounds of Hackathon spanning over four months and leading up to final presentation amidst COVID-19. Building a Machine Learning pipeline to pinpoint FMCG's end-to-end efficiency. Combining historical sales on items with similar names, brands, sales channels and ingredients to predict new items' sales volume. Embedding text data with PERT model with appropriate dimension reduction as input to CatBoost model. Achieved second highest test accuracy amongst contestants and won fifth place overall.
Unilever Hackathon: New Product Sales Volume Estimation
