About

Your time is limited, so don’t waste it living someone else’s life.

Hello there, welcome to my site, this is Donghan (Adrian) Liu, a data enthusiast, problem solver, curiousity explorer, impact maker.

About ME

My interests range from Machine/Deep Learning algorithms (theoretical/application), Data Science techniques to impactful Data-Driven Decision in the real-world. Love implementation and problems-solving from life, society and business.

I chose this field of study because I’ve always been interested and stay curious in data science. Data science is the magic tool that you are using to make the real impact to our society and business, which is the way that I admire and would like to use it as a tool achieve my personal value. Also, I’d really describe myself as data enthusiast, I enjoy and passionate working on a data science projects, so in the past three years, I participated some data competitions and Microsoft hackathon, where I ever won the first place and couple of times in top 5, which brings me a great hands-on practice on data modeling and tuning skills. Also, while I was undergrad majoring in statistic, I learned how to write statistical report and interpret the results to non-expert person by solving the problem that was related to data analysis.

You can reach me at LinkedIn, Email

[GitHub]

Education

• I earned my Master’s degree of Information Studies at McGill University, May 2020, with high honor.

• I also gained my Bachelor’s degress of Science in Statistics at University of Illinois Urbana-Champaign, May 2018, with high honor. Fighting ILLINI

Research

Former research assistant of Data Mining and Security (DMaS) Lab, worked in Mouse Stress Determination project using Deep Learning techniques.

• Technical:
  • Among the first to propose the deep learning method in mouse hippocampus stress determination.
  • Process time-seris data: LSTM/GRU with PyTorch/Tensorflow framework
  • Process videos/images: OpenCV package
  • Demonstrate the group of neurons that spike together/similar influence to motor activity: Statistical correlation/ Association rules
  • Model hyper-parameters tuning, performance evaluation
  • Data visualization and presentation
• Teamwork/Leadership:
  • collaborate with five research assistants with various academic background
  • ensure the completion of research project in a timely manner by weekly updates and active communication
  • efforts/outputs had been praised by supervisor

Teaching

• Teaching Assistant for Methods of Applied Statistics at University of Illinois Urbana-Champaign, January 2018 - May 2018

• Teaching Assistant for Computer Programming at McGill University, August 2019 – May 2020

Machine Learning Related Project Experience

• Binary Classification - Time-seris Data Modeling, January 2019-May 2020 - Python, PyTorch, Long-Short-Term-Memory, Gated-Recurrent-Unit, Transformer, OpenCV [source code]
• Images Classification - Implementation of Multi-Layer Perceptron by Numpy and Convoluntional Neuron Nets, May 2020 - Python, PyTorch, Relu, Sigmod, Leaky_relu, Softmax, CNN [source code]
• News/IMDB reviews Classification - Natural Language Processing, April 2020 - Python, Model Tuning and Comparison, Preprocessing (nltk), Random Forest, SVM, KNN, Adaboost, Naive Bayes [source code]
• Implementation of Gradient Descent and Naive Bayes by Numpy - Multiple Data Sources, March 2020 - Python, Logistics, Naive Bayes [source code]
• Association Rules for Grocery Purchase Records, December 2018 - Python, Association Rules [source code]
• Analysis Of Income Inequality In The US, May 2018 - R, Linear, Stepwise Model Selection, Polynomial, KNN, Random Forest, Ridge, Lasso, CV [source code]
• Flights Status Comparison, May 2018 - R, Hive, CMD, Treemap, Heatmap, Stream graph, Logistics [source code]
• Statistical Methods Applications in Financial Engineering - Variance Reduction, December 2017 - R, Importance Sampling, Stratified Importance Sampling, Control Variates, Antithetic Variates [source code]
• Health Model Development, May 2017 - R, Log-transformation, Box-cox-transformation, Residual and Normality Check, Cook’s Distance/Outliers, AIC, BIC, Stepwise Model Selection [source code]

Data Science Competitions

• Finalist at Microsoft Hackathon - Build Business Model, December 2018 - Azure, HTML, Probabilistic model [source code]
• ANZ Bank Data Competition - Top Ten of Eighty-two Teams, November 2018 - R, Imbalanced Data, Correlation, Importance, Random Forest, Gradient Boosting, Oversampling [source code]
• Synchrony Datathon - Top 5, R, Random Forest [source code]
• Sunrise Futures LLC Data Competition - Winner First Place, April 2017 - Python [source code]

Association Experiences

• Data Researcher at Events@UI Studio, November 2016 - January 2018
• Consultant & Analyst at Statistics in the Community, August 2017 - January 2018

Award

• Dean’s list 2016 – 2017, 2017 – 2018, University of Illinois Urbana-Champaign

• Graduate Program Scholarship recipient 2018 - 2019 , McGill University

• Minor/Certificate of Computational Science and Engineering July 2019, University of Illinois Urbana-Champaign

Media

Intuitional Meaning of WeChat Emoji - Social Research