Class of 2021

Photo of Ming Zhang

Ming Zhang graduated from UC Santa Barbara in June 2020 with both BS and MS in Actuarial Science (Combined Program). He found his interest in data science during his second year in a machine learning class, where he used several statistical models to predict the result of presidency election in 2016 by using the previous polling information. 

After Ming realized the magic power of data science, he took many classes about statistical models and programming, and actively sought research/internship opportunities in this area. In the Times Series class, he tried to predict the car plates prices in Shanghai for the next few months; in the Statistical Data Science class, he built a text mining model to evaluate a local restaurant through Yelp Reviews. Meanwhile, Ming participated an actuarial research sponsored by California State Association of Counties – Access Insurance Authority. After comprehending and cleaning the data, he used machine learning techniques like Logistic Regression, Support Vector Machine and Random Forest to build a model, which helped his sponsors to predict cases that could be “Compromise & Release”. Moreover, Ming also tried to improve himself through working as a data scientist intern at BizSeer Technology, and he was asked to construct a short-term times series prediction model for transaction numbers in a bank. In the beginning, he tired many complicated models such as LSTM, SARIMA and Gaussian Process Regression, but none of these worked well due to the high frequency of the data and short running time requirement from his clients. After numerous different trials, he found there were strong linear relationships between the differences of transaction numbers at the same time on different days, and he came up with a model prototype based on linear regression. With more experiments conducted, he decided to use combination of different linear models to lower the model bias, and finally his model could predict transaction numbers for every minute in the next 6 hours. This internship granted him a lot of experiences about how to apply the theoretical idea into the data science world, and he realized that he should improve his programming skills and accumulate more algorithms knowledge in the future study.

As a candidate for the MSiA program at Northwestern University, Ming is thrilled to receive professional training in Analytics. With the preparation of various courses in this program, he believes that he can be a real data scientist who can have a profound mathematical background as well as a strong business acumen.