Home

Education

---

Skills

• Software: Snowflake SQL, MS SQL, Azure, Tableau and Mode Analytics.
• Programming: SQL, Python, R, Java and C++; PowerShell, Zsh, Bash and Git.
• Field Knowledge: Data Quality Assurance, Data Migration, Data Engineering, Electronic Medical System, Machine Learning, Data Structure and Algorithms.

---

Experience

---

Activities

• Implemented and picked 10+ trend, momentum, volatility, and volume indicators. Constructed a signal system based on neural network models. The beta version provided 0.65+ Sharpe Rate and 80%+ non-leverage APY in the backtest.

• Provided basic introduction to financial investments.
• Produced educational content for risk management, trading psychology, and trading strategy.
• Illustrated the plan and execution of strategies.

• A crypto market version of traditional candlestick patterns, with Chinese language option.
• Improved the original *All Candlestick Patterns* indicator in signal illustration.
• Added more options for adjusting recognization's precision.

• Distplyr: provides a grammar for manipulating the probability distributions under a clean and user-friendly framework.
• Distionary: provides distribution types as dst_*() family of functions and provides the basic objects to be manipulated under distplyr.
• Presented and introduced the two packages to public.

• Cleaned up and preprocessing data with 400k+ entries from Kaggle's competition "Quora Question Pairs".
• Developed an 85%-accuracy LSTM with Word2Vec model for question pair classification based on sentence similarity.
• Researched on 1D-CNN embedding layer with class weights, TFIDF, and mover distance etc., 86% accuracy achieved.

• Exploratory data analysis and time series processing.
• Developed a Bayesian structural time series model for seasonal prediction on S&P 500 index with 84% accuracy.
• Built Gaussian process models in python for monthly prediction on S&P 500 index.

• Applied the survival analysis on the conversion rate of first purchasing of new users.
• Constructed a Cox PH model and an Aalen's Additive regression model to recognize the vital factors on users' behavior.
• Provided marketing strategy based on the key reported factors for specific targeted customers and markets.

• Collaborated in team and developed a longitudinal study focusing on the patients' evaluation of surgery outcomes.
• Drafted exploratory analysis and univariate analysis, using in-depth visualization and sufficient A/B tests.
• Preprocessed and restructured raw data with 1.5k+ entries for logistic regression and linear mixed effect models.

• Presented exploratory data analysis and missing values analysis on the hospital measurements of HCV patients.
• Built OvR/MvM multiclass logistic-regression-based classifiers with 95.4/96.4% accuracy for the patient recognition.
• Implemented a Naive Bayes Classifier with 93.81% accuracy as the benchmark for the multiclass classifier.

• A photographic project of a series with diagonal layout.
• Explore the collision between building materials and the sky.
• Explore the intersection between geometric structures and natural views.

• Completed the challenge of traffic analysis from Mobikit, for improving the safety function in auto driving.
• Conducted modification on nominal data, data cleaning and missing values analysis.
• Constructed logistic regression modelling and performed model selection.

• Data modification, feature selection, exploratory data analysis and multiple hypothesis test with high power.
• Presented insights of the data and interpreted data visualization for non-academic audience.
• Offered specific suggestions regarding daily lifestyles of university athletes.

• Tested complied codes and maintained coding consistence.
• Designed an educational language based on JavaScript which could be compiled to low level subsets of JS and C.
• Kept the language as strictly typed, multi-paradigm, and compiled with modern and clean style.