Elucidate geological hazards in the Dien Bien - Lai Chau fault zone, assessing the geological hazards (earthquake, landslide) with the support of machine learning, artificial intelligence.

Theoretical results: By analyzing remote sensing, geological-tectonic characteristics, topography, and geomorphology data in the Dien Bien - Lai Chau fault zone (ĐB-LC) and surrounding areas. The project has identified the seismic fault segments affecting the study area. The earthquake hazard map has been established according to two methods of probability and determinism on a scale of 1:250,000. With the support of machine learning and artificial intelligence, the project has successfully built an instruction and map to predict landslide susceptibility map at the 1:50,000 scale in Muong Lay and Sin Ho with high accuracy (estimated accuracy according to ACC = 92.86%, AUC = 0.982). High to very high landslide susceptibility class is predicted to be concentrated in the area along the main fault zones, especially along the ĐB-LC fault zone and next to the main roads. The BFPA model gives the highest prediction results for the study area, and it is recommended to apply to other areas with similar conditions.
Applied results: The results of the project are applicable in planning and construction to prevent natural disasters caused by earthquakes and landslides in the study area.

The results of the project initially evaluate in detail the sources of seismic faults and use the new earthquake attenuation model in the earthquake hazard assessment in the ĐB-LC fault zone and surrounding areas. The project has built a detailed process of landslide hazard prediction using machine learning, and artificial intelligence and for the first time earthquake hazard factors are considered in landslide assessment in the ĐB-LC fault zone and surrounding.

- Scientific papers in referred journals (list):
SCI-E Index:
1. Binh Thai Pham, Tran Van Phong, Trung Nguyen-Thoi, Phan Trong Trinh, Quoc Cuong Tran, Lanh Si Ho, Sushant K. Singh, Tran Thi Thanh Duyen, Loan Thi Nguyen, Huy Quang Le, Hiep Van Le, Nguyen Thi Bich Hanh, Nguyen Kim Quoc, Indra Prakash, 2020. GIS-based ensemble soft computing models for landslide susceptibility mapping. Advances in Space Research 66 (2020) 1303–1320.
2. Nguyen Van Dung, Nguyen Hieu, Tran Van Phong, Mahdis Amiri, Romulus Costache, Nadhir Al-Ansari, Indra Prakash, Hiep Van Le, Hanh Bich Thi Nguyen, Binh Thai Pham, 2021. Exploring novel hybrid soft computing models for landslide susceptibility mapping in Son La hydropower reservoir basin. Geomatics, Natural Hazards and Risk 2021, Vol. 12, No. 1, 1688–1714.
ESCI/Scopus/VAST01 Index:
1. Tran Van Phong, Hai-Bang Ly, Phan Trong Trinh, Indra Prakash, Dao Trung Hoan, 2020. Landslide susceptibility mapping using Forest by Penalizing Attributes (FPA) algorithm based machine learning approach. Vietnam Journal of Earth Sciences 42 (3): 237-246.
Chapter:
1. Tran Van Phong, Nguyen Duc Dam, Phan Trong Trinh, Nguyen Van Dung, Nguyen Hieu, Cuong Quoc Tran, Tung Duc Van, Quan Cong Nguyen, Indra Prakash, Binh Thai Pham, 2022. GIS-Based Logistic Regression Application for Landslide Susceptibility Mapping in Son La Hydropower Reservoir Basin. Published in: CIGOS 2021, Emerging Technologies and Applications for Green Infrastructure. Lecture Notes in Civil Engineering. Springer publisher.
- Technological products (describe in detail: technical characteristics, place):
1. Map of active faults and seismic faults, concentrated for the key area (1:250,000 scale in Arcmap software format, save at Institute of Geological Sciences (VAST))
2. Map of Earthquake hazards (1:250,000 scale, PGA maps with probability exceeding 10%, 5%, 2%, 1%, 0.5% in 50 years and is deterministic (Arcmap software format, save at Institute of Geological Sciences (VAST)))
3. Landslide hazard map based on applying machine learning algorithms, artificial intelligence, key area selection (1:50,000 scale in Muong Lay (Dien Bien) and Sin Ho (Lai Chau), Arcmap software format, save at Institute of Geological Sciences (VAST))

The research results of the project can be applied in practice, as a reference in teaching at universities.