Dongqi Pu

M.Sc. student in Data Science
The University of Queensland(UQ)

Currently, I'm pursuing a Master's degree in Data Science at University of Queensland. Before coming to UQ, I got my bachelor degree from School of Computer Science at Central China Normal University, Wuhan.

My research works mainly focus on the Machine Learning and Deep Learning.

This is My website and please feel free to download my CV / Résumé .

Education

M.Sc. in Data Science
The University of Queensland, Australia
Feb.2020 - Nov.2021(Expected)

B.Eng. in Software Engineering
Central China Normal University, China
Sep.2015 - Jul.2019
Major GPA: 86.48/100
Ranking: Top 1(Junior), Top 3(Sophomore)
Outstanding Graduate, 2019

Exchange Student. in Computer Science
‎National Tsing Hua University, Taiwan
Sep.2018 - Jan.2019
Supervisor: Prof.Liang-BiChen

Research & Project

The Application Research of Honey Encryption in Protecting Privacy Data
Supervised by Prof.GUO Yajun , December.2018----May.2019
< Graduation Project>, Dongqi Pu

• Base on password encryption, we introduce the concept of the distribution-transforming encoder. The system will output a plausible-looking decoy message to confuse the attacker when decrypting a message with exhaustive-key-search method. In this project, we apply honey encryption to protect the users’ digital wallet, which can solve the drawback of weak passwords using in the users’ bank card accounts and their PIN number. The result shows that the security of honey encryption is higher than the existing password-based encryption, and the decoy message is also too difficult to distinguish.

"The Beauty of Math" —— A science popularization website(Team Leader)
Supervised by Prof.Shuangjie Peng , April.2018----April.2019
< Dean of the School of Mathematics and Statistics, CCNU >
Zihan Chen, Dongqi Pu, Zirui Li, Kaiqi Xu, Ya Zhang  

• In university, mathematics is a fundamental subject, whether you are a science and engineering major or an economics management student. But mathematics classes focus mainly on teaching higher mathematics knowledge and its applications, with mathematical knowledge seen only as a tool for learning follow-up courses or solving scientific and technological problems. Mathematics in the ideological, spiritual and human sense is mostly ignored. Thus our team intends to build a website to increase the common public interest in math for its own sake.

Short-term prediction of commercial housing price in Hongshan district based on Web search data (Project Leader)
Supervised by Prof.Fuzhe Zhao , March,2017----March,2018
Dongqi Pu, Dingling Shao, Tingyu Wu, Jing Zeng and Wenying Qian   [Github]

• Network retrieval data contains useful details regarding customers interests, and are representative of expectations and likely economic behaviors. Therefore, Network retrieval tracing and real estate keyword search can be used to formulate short-term housing price predictions. Specifically, using Baidu Search Index (BSI) as sample data, we disposed of default values, filtered out zero-variance variables and selected 23 key variables from a total of 60 search words. To get the best models, we then fitted and forecasted commercial housing prices in Hongshan District, Wuhan City by using six regression models ( Multiple Linear Regression, Classification and Regression Trees, Bagging, Random Forest, Support Vector Machine) in the empirical research section. The results show that the Linear Regression Model and Random Forest Model delivered the best performance, with an average error rate of -0.11% and 0.13% respectively.

Publications

Privacy Data Protection Based on the Honey Encryption
GUO Yajun(Advisor), PU Dongqi
NETINFO SECURITY(Accpeted)
• The security of honey encryption is higher than the existing password-based encryption, and the decoy message is also too difficult to distinguish.

Housing Price Prediction Based on Web Search Data
Dongqi Pu, Dingling Shao, Tingyu Wu, Jing Zeng and Wenying Qian
Information & Communications pp.17-19, January.2018.
• Utilized six regression models in the empirical research section and The Linear Regression Model and Random Forest Model delivered the best performance, with an average error rate of -1.1% and 1.3% respectively.
• This paper has been quoted two times by now.

Application of Data Mining in Artificial Intelligence
Dongqi Pu
China Computer & Communication pp.157-159, October.2016.
• A brief introduces to artificial intelligence and data mining applications.
• This paper has been quoted four times by now.

Decadent Time
Sole Author
Published by Chinese Press (HongKong)
• ISBN: 978-988-15209-0-6.
• There were 35000 volumes of sales.

Selected Awards

  • Outstanding Graduate, 2019
  • Overseas Exchange Scholarship, 2019
  • Science and Innovation Star, 2017
  • There were only ten students received this honors in the university in a year.
  • Boya Scholarship, 2017 & 2018
  • Honor Roll student, 2017 & 2018
  • Shuren Scholarship, 2016

Failures

Most of what I try fails, but these failures are often invisible, while the successes are visible. So I am trying to record those of the bad things and encourage myself.
  • Degree programs I did not get into
    2018       PhD Program in Computer Science, University of Manchester
    2015       BSc in Computer Science, Sun Yat-sen University
  • Academic positions and fellowships I did not get
    2018       Summer Research Intern, Tongji University
  • Work positions I did not get
    2018       Data Mining Internship, Alibaba Group
  • Scholarship I did not get
    2018       National Scholarship

News