2021 NBA Champion Predictor (Cont.)
June 2021
- Used the same dataset as the previous prediction to continue exploration of the NBA 2021 season statistics.
- Train test split the data using scikit learn.
- Same core statistics were used as the previous NBA champion prediction.
- Used tensorflow with different layers and techniques such as Batch Normalizatoin, Dropout, Early Stopping.
- Fitted the 2021 season data into the model, Brooklyn Nets were predicted to be 2021 NBA champion, with Milwaukee Bucks insanely close behind.
- With the elimination of the Brooklyn Nets, Milwaukee Bucks are now the top of the list from this prediction to become the 2021 NBA Champion.
July 2021
- Milwaukee Bucks is the 2021 NBA Champion
2021 NBA Champion Predictor
May 2021
- Retrieved NBA team datasets since 1980 season to May 1st 2021 from Kaggle.
- Cleaned and joined the data using tidyverse package in R.
- Created a decision tree with Gini criterion using scikit learn in python using past NBA seasons’ data with champion labelled.
- Included factors such as field goal percentage, 3 point percentage, free throw percentage, win rate, turnovers, and assists.
- Trained, tested, and predicted the NBA champion for the 2021 season 1000 times.
- The Milwaukee Bucks are predicted to be the 2021 NBA champion.
- I usually support the Raptors but during this NBA playoffs I will be cheering on for the Milwaukee Bucks and Giannis.
Shrager
February 2021
Winner of GM – Putting EVerybodyIn an EV prize at UofTHacks VIII Hackathon.
- Web application that connects EV owners, offering a more convenient way to find charging for your EV or rent out your charger!
- Collaborated on a web application that connects EV owners, allowing them to rent out their chargers or let them look for a nearby charger to charger their electric car.
- Worked on the back-end using express.js, searching using Google Maps API and database using MongoDB.
LunaBot
March 2020 – Present
Multifunctional Discord bot using Node.js to perform various functions in private discord server such as
- Joining voice channel to play music
- Ranking/exp/class systems through chatting
- Daily update of COVID-19 cases in Ontario
- Uses RIOT API to provide League of Legends information
- Uses iex stocks API to provide stock quotes and information
- Utilizes MongoDB for storing all user information
- Creates angry reacts ranking based on user reacts on each messages
UofT Help Bot
July 2020 – August 2020
Discord bot using Node.js & Nikel API to allow users to search up information regarding courses, exam schedules, and more.
Currently serving 4,300 discord users.
Tech Conference Application
September 2020 – December 2020
Tech Conference Application is a project for CSC207 Software Design course at University of Toronto during the Fall 2020 Semester. The application allows users to sign up for accounts as different user types such as Organizer, Speaker, User, VIP. Each user types contain different functions of their own. Organizers are able to create and manage events and rooms, speakers are able to speak at different events, users and VIPs are able to sign up at different events. Every user has the ability to message and add other users to their contacts. The data is stored locally and PDF output of the event schedule and room is available.
- Contributed towards the event functionality and program compilation and assisted on the controllers.
- Applied clean architecture rules and design patterns towards the design of the application.
- Utilized Git to version control with group members, created JavaDoc for documentation.
Anime & Manga Trend Analysis
July 2020 – October 2020
- Used Node.js and Kitsu.io API to create a script and collect data from 18,000 animes since 1907.
- Set up a Python script using BeautifulSoup and Selenium WebDriver that scrapes 23,000 mangas and their information from Mangakakalot.
- Utilized pandas to clean the data and ltered based on dierent attributes.
- Displayed trends in categories with graphs and bar chart race animation using Matplotlib and plotly.
Mall Monitor
January 2021
Awarded as finalist for Hack The North 2020++ Hackathon.
- COVID-19 tracking system for malls, allowing users to track the amount of people in and out of their store
- Worked on a product that detects people in and out of a specic area using camera footage with YOLOv3 and OpenCV.
- Contributed to the back-end with express.js and database with MongoDB, assisted on Microsoft Azure deployment.
GameConnect
October 2020
Social media website that allows people who play games to connect and interact with each other. Contains functions such as adding games to your game list, post, like, and comment.
- Used Django, Python, HTML, and CSS to create a social media website for people who play games.
- Allows user to sign up and modify personal information such as country, bio, name.
- Created posts section for user to post, like and comment, allowing users to interact with each other.
- Contains a list of popular games which user can browse and add to display on prole.
Header image taken at Gairloch Gardens, Oakville on May 16th, 2019