Leading AI and Data Science projects. Providing consulting to organisations and companies:NSW Department of Education ITD, NSW Department of Education EDConnect , NSW Health Murrumbidgee Local Health District, Me Squared, MoneySoft, etc

1. Prediction on In-hospital patient readmissions with MLHDo Based on patient demographic, health history, and hospital admission information, I developed a predictive model using decision tree (with R) to target the group of patients at high risk of readmission with accuracy of 70%. This will help the hospital to perform proper intervention on these patients to reduce their chance of readmission. o The predictive result is updated every day and displayed in PowerBI for easy use to doctors and clinicians. 2. Exploratory analysis and prediction on hospital visits with respiratory issue with MLHDo Performed exploratory analysis on the data set to discover the relationship between hospital visits with respiratory issues and factors of patient health condition, weather changes, seasons, and air quality. o Developed negative binomial regression model to predict how many patients will visit hospital with respiratory issue each day.o Developed logistic regression model to predict the risk of hospital visiting with respiratory issue of a potential patient list each day. 3. Prediction on Out-of-stock situation with SalesLinko By analyzing the history out-of-stock data, I developed a time series model to forecast out-of-stock situation for a supermarket supplier, to prevent the loss of sales due to insufficient stock4. Exploratory analysis on transaction records and customer purchase patterns with Humanitio Transaction records from different source such as bank account and superannuation account contain huge valuable information about customers for marketing usage. I performed a variety of exploratory analysis (using text matching and clustering techniques in R) on transaction data, to dig out the assets. o Categorizing transaction records to find the purchase habit of customers. o Extracting location information from transaction records. o Analyzing income transactions to check customer employment status and investment status

Developing researchers grant project and publication interactions network for Hawkesbury Institute of Environment with R Shiny.

Developed an algorithm to analyze human tissue images, performing predictive model analysis on the data sets to determine whether the tissue is normal or in disease (used SAS, Matlab)

Observed and taught some of the sections of two undergraduate entry level mathematics courses for two quarters. Learned how to arrange course contents to make it clear and easy for students to learn, how to inspire students’ interest and guide them to fulfill the learning goal.

Analyzed large scale data sets (over half million observations each data set) in my PhD research project: gathering data generated from different DNA microarray image analyzing methods; building predictive models to compare and evaluate the performance of each image analyzing algorithm (used R, Matlab, Java)Built statistical models in logistic regression and linear regression; performed hypothesis testing, quantitative analysis, multivariate statistics, and experimental design & analysis (used R, SAS)Implemented and developed the image analyzing numerical algorithms into executable user-friendly programs (used Java, Matlab); Developed two Android Phone apps in my computer science courses: Chemical lab flame test, and Piano play game (used C++)

Chief Editor of an E-magazine that reports campus news for students readers.Made connections between Teaching college with other colleges in Qingdao University.

Internship as a mathematics teacher for an 8th grade class of 46 students.