Project's information

Project's title Study on Extended Rough Set based Feature Reduction Methods and Application in Objects Classification in Satellite Images
Project’s code VAST01.10/20-21
Research hosting institution Institute of Information Technology
Project leader’s name Assoc.Prof. Nguyen Long Giang
Project duration 01/01/2020 - 31/12/2021
Project’s budget 600 million VND
Classify Excellent
Goal and objectives of the project

- Propose tolerance based attribute reduction methods in incomplete decision tables by filter-wrapper approach to reduce the number of attributes and improve the efficiency of classification model.
- Applying proposed attribute reduction methods in solving attribute reduction in satellite images classification.

Main results

Theoretical results:
- Propose a distance based filter-wrapper attribute reduction algorithm in imcomplete decision tables.
- Propose distance based incremental filter-wrapper attribute reduction algorithms in  dynamic imcomplete decision tables in the cases of adding or removing an attribute set;  adding or removing an object set.
- Training 01 PhD student, 01 MsC student with thesis topics in the research direction of the project.
- 03 papers in the research direction of the project. In which 02 paper in the SCIE journal; 01 paper in the Proceeding of  National Conference on Information Technology and Telecommunication, 2020.    
Applied results:
- Develop software for satellite images classification (two classes: 1 - with ships;   0 - without ships) by using proposed attribute reduction algorithms.
- Experiment and evaluation of software implementation on satellite image data sets from VNREDsat-1 and others data resources.

Novelty and actuality and scientific meaningfulness of the results

- Propose a distance based filter-wrapper attribute reduction algorithm  IDS_FW_DAR in imcomplete decision tables.
- Propose incremental filter-wrapper attribute reduction algorithms in dynamic imcomplete decision tables by using proposed distance: algorithms IDS_IFW_AO, IDS_IFW_DO in the cases of adding or deleting an object set;  algorithms IDS_IFW_AA, IDS_IFW_DA in the cases of adding or deleting an attribute set.
The project applies the proposed algorithms to reduce attributes in the satellite images image classification problem (into 02 classes: 1 - with ships; 0 - without ships) to improve the efficiency of the classification model.

Products of the project

Scientific papers in referred journals (list):
02 paper in the SCIE journal, 01 paper in the Proceeding of  National Conference on Information Technology and Telecommunication.

  • Nguyen Truong Thang, Nguyen Long Giang, Hoang Viet Long, Nguyen Anh Tuan, Tran Manh Tuan, Ngo Duy Tan, “Efficient Algorithms for Dynamic Incomplete Decision Systems”, International Journal of Data Warehousing and Mining, Volume 17, Issue 3, pp. 44-67 (SCIE), 2021.
  • Nguyen Long Giang, Le Hoang Son, Nguyen Anh Tuan, Tran Thi Ngan, Nguyen Nhu Son, Nguyen Truong Thang (2021), “Filter-Wrapper Incremental Algorithms for Finding Reduct in Incomplete Decision Systems when Adding and Deleting an Attribute Set”, International Journal of Data Warehousing and Mining,  Volume 17, Issue 2, pp. 39-62, (SCIE), 2021.
  • Nguyen Anh Tuan, Nguyen Van Thien, Nguyễn Long Giang, “About incremental filter-wrapper attribute reduction algorithms in imcomplete decision tables when adding, deleting an attribute set, Proceeding of  National Conference on Information Technology and Telecommunication, Quang Ninh, 5-6/11/2020, pp. 477-482.

Technological products (describe in details: technical characteristics, place):
- Report on tolerance rough set based filter-wrapper attribute reduction methods in imcomplete decision tables.
- Report on tolerance rough set based incremental filter-wrapper attribute reduction methods in dynamic imcomplete decision tables.
- Experimental program for satellite images classification by applying proposed attribute reduction methods.
- Report  the experimental results of the program.
- Report on the results of the project
- Scientific papers (evidences in the report of the project )
- Training support 01 PhD student, 01 MsC student (evidenced in the report of the project)

Research area

Proposed implementation at the Space Technology Institute, VAST.

Images of project
1655266673136-38. nlg.png