Lorentz Yeung
Worked with Microsoft as Data Analyst / AI Certificate (Harvard University) / MSc in Digital Marketing / GC in Art History / Bachelor in Economics /
Back To Front PageAI Certificate from Harvard University
May 2021
It was awarded by the HarvardX of Harvard University. Topics covered are:
probability theory, Bayesian networks, Markov models, machine learning, neural networks, and natural language processing.
The view link below goes to: https://cs50.harvard.edu/certificates/
Ranked 7 in a Kaggle Community Competition
June 2022
My submission of a classification model finally ranked 7th in the competition. I have another late submission scored even higher and is on par with the first runner up.
Built a serverless ML model
July 2022
A serverless deployment of my machine learning model with PyScript. The model was built on Xgboost, but is finally deployed with Logistic Regression, and Gradient Boost because PyScript doesn't support Xgboost. Please click the button below to test it out. It takes about a few seconds to a minute to load, thanks a lot for the patience.
Automated trade signal generation model with daily web-scraping and sentiment analysis for Algo-trade
November 2021 til now
A complicated Algo-trade project with web scraping, server creation, data fetching, sentiment analysis, model creation, server maintenance and risk management, subscriber external database update, sending out daily trade-signal notification emails, etc. You are welcome to sign up for my automated email notification below. It is a friendly sharing, totally free of charge.
Generative Stable Diffusion model with modern UI
October 2022
I developed a web app with generative Stable Diffusion model and tkinter. It allows you to type in a prompt into the text box and have an image generated. The avatar above is Goku plus Ironman.
Automated monthly London property price analysis
March 2021 til now
The reports are for the HongKonger community, and the process is totally automated with Python. Processes included but no limited to are, web scraping, data engineering, data visualization, insights generation, and PDF report generation. Uptil the end of 2023, there had been 60k reader count of the report series. The automation vastly reduced the entire analysis time from 24 hours to 30 mins only; that is 98% reduction in time.