Jingyuan Zhang

Master Student

OSCAR Lab
School of Software

Dalian University of Technology

Dalian, Liaoning, P.R. China, 116621

Tel:  +86-135 9135 9643
Email:  jyzhang.dut at gmail.com


Research Interests      Brief Biography     Research Work      Publications        Lab Presentations         Teaching Assistant

Research Interests

  • Data Mining:  I focus on developing alternative clustering algorithms for fast-evolving stream data. I'm also interested in areas of uncertain data mining and ant colony clustering.

  • Heuristic Search:  I focus on applying Ant Colony Optimization algorithms to problems arising in software requirement engineering.

Brief Biography  

I'm a third year Master student in School of Software, Dalian University of Technology (DUT). I work in OSCAR Lab (Optimizing Software by Computation from ARtificial intelligence). My advisor is Dr. He Jiang. I obtained my Bachelor degree in Software Engineering (Sept. 2005 - Jul. 2009) in DUT.

My Curriculum Vitae is here [CV]. 

Research Work

  • Alternative Clustering on Stream Data: We issue the new problem of alternative stream clustering, which aims to find two high quality and dissimilar macro-clusterings in a given data stream. To tackle this new task, we propose a new algorithm named AltStream consisting of two components.
    • The online component simultaneously maintains two alternative groups of micro-clusters.

    • A new measure SOBD (Sum of Balanced Distance) approximately evaluates the dissimilarity between two clusterings.

    • The offline component extracts some semi-supervised information tocreate two alternative macro-clusterings.

  • A Hybrid ACO Algorithm for the Next Release Problem: Next Release Problem (NRP) is a NP-hard problem in requirement engineering. It seeks to to balance customer requests, resource constraints, and requirement dependencies by requirement selection. To tackle large NRP instances, we propose a new algorithm named Hybrid ACO (HACO).
    • Multiple artificial ants take both pheromone trails and neighborhood information to construct initial solutions.

    • HACO incorporates a local search FHC (First-found Hill Climbing) to improve solution qualities.

    • HACO employs an automatic tuning procedure CALIBRA to effectively determine the interrelated parameter settings.

    For more information about NRP, please refer to the supplement materials presented by Dr. Jifeng Xuan in OSCAR Lab.

Publications

  • He Jiang, Jingyuan Zhang, Shichao Qu, "Alternative Stream Clustering", under review in SDM Conference, 2012.

  • He Jiang, Jingyuan Zhang, Jifeng Xuan, Zhilei Ren, Yan Hu, "A Hybrid ACO Algorithm for the Next Release Problem", in Proceedings of 2nd International Conference on Software Engineering and Data Mining (SEDM 2010), Chengdu, China. June 23-25, 2010, pp. 166-171. [PDF], [IEEE]

Lab Presentations

    The following is about my seminar presentations to OSCAR members. If you are interested in any of them, please contact me by email for a copy.

  • 2011-07-29    AltStream, HPStream and DenStream: It's about three algorithms for stream clustering. AltStream aims to find two results while HPStream and DenStream only return on single clustering. [Abstract]
  • 2011-04-21    A Framework for Alternative Clustering on Stream Data: It's an initial framework for alternative stream clustering. Some improvements were made later. [Abstract]
  • 2010-11-30    Alternative Clustering on Stream Data: The feasibility of finding alternative clustering on stream data was analyzed here with some real-world examples. The methods to compare two clusterings were also presented. [Abstract]
  • 2010-09-20    Finding Alternative Clustering: It's about four alternative clustering techniques on traditional datasets, including COALA, Multi-view, ADFT and Constraint optimization. [Abstract]

  • 2010-06-18    Uncertain Data Mining: It's about a literature investigation of uncertain data mining, including clustering, classification, frequent item set mining, and outlier detection. [Abstract]

  • 2010-06-04    Semi-Supervised Clustering with Constraints: It's about constraint clustering, including COP-kmeans, distance metric learning, EM and HMRF Model. [Abstract]

  • 2010-01-08    A Hybrid ACO Algorithm for the Next Release Problem: It's about the ACO algorithm for NRP problem with the parameter tuning procedure CALIBRA. [Abstract]

Teaching Assistant

    I have been a teaching assistant since fall 2008. Here is a list.

  • Fall 2011          Discrete Mathematics (Instrutor: Yan Hu)
  • Fall 2010          Computer Organization and Structures (Instrutor: Jie Wang)
  • Fall 2009          Java Language Programming (Instrutor: Yan Hu)
  • Spring 2009     Operating System Principle (Instrutor: Yan Hu)

  • Fall 2008          UML (Instrutor: Zheng Wang)


Last Modified: 31/10/2011