KEVIN YANG
[ addcib@r.postjobfree.com Ó 226-***-**** Kevin Yang
EDUCATION
Statistics and Economics Majors
University of Toronto
Sep 2017 – Apr 2021 (Expected)
SKILLS
Languages
Python R Java
JavaScript SQL
Technologies
MS Office NumPy Matplotlib
Seaborn Pandas Plotly
SVN Github
ML Techniques
Linear/Logistic Regression
K Nearest Neighbors
Decision Trees K Means Clustering
Cloud Technologies
Azure AWS
EXTRACURRICULAR
Event Executive
Canadian Asian Student Society
Sep 2019 – Present
Executed events to bring students
together and create a community on
campus
COURSEWORK
Relational Databases and Web Dev:
SQL to query data
HTML/CSS/JS to build a website
Stochastic Processes:
Markov Chains
Gaussian Processes
Brownian Motion
EXPERIENCE
Product Data Analyst
Humi
May 2019 – Aug 2019 Toronto, ON
Designed reports tracking KPI health (Net MRR Churn Rate, NPS, Retention Rates) This enabled the Client Experience team to iden- tify engagement trends and prevent churn
Implemented Python scripts on top of SQL to effectively query data from the Humi database in any format and layout required
Created a Net MRR Churn deck to show members of the Client Experience team their core and at-risk clients. This displayed each member’s monthly portfolio growth/decline.
Discovered a discrepancy between NPS scores given by users and power users, this initiative was proactive churn prevention and was credited with addressing 50% of all churn. This tool enabled churn to be minimized
PROJECTS
Analysis of Bank Stocks - Financial Crisis 2016
May 2020
Collected, pre-processed, and analyzed large sets of Financial Cri- sis data obtained through data mining the web
Designed cluster heatmaps to outline correlation between Bank stock closing prices, displaying market behaviour patterns
Leveraged line plots to showcase market trends for the desired time period, utilizing close price data. This presents a huge crash for Citibank in 2009 due to the Obama Inauguration. Tools: Pandas Data reader, Pandas, Seaborn, Numpy, Cufflinks Optimizing Transportation in Populated Areas - ML Approach
Present
Trained machine learning models using live-feed street intersec- tion video to successfully implement object detection, pathing, and human count for storefront traffic
Designed and executed an informative platform for the average resident/tourist. With a motivation from the user, the platform recommends hot spots in any area of interest
Tools: OpenCv, Keras, Tensorflow