首页|Researchers from CSIR -Central Building Research Institute Describe Findings in Machine Learning (A Comparative Evaluation of Statistical and Machine Learning Approaches for Debris Flow Susceptibility Zonation Mapping In the Indian Himalay as)
Researchers from CSIR -Central Building Research Institute Describe Findings in Machine Learning (A Comparative Evaluation of Statistical and Machine Learning Approaches for Debris Flow Susceptibility Zonation Mapping In the Indian Himalay as)
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
Investigators publish new report on Ma chine Learning. According to news reporting out of Roorkee, India, by NewsRx edi tors, research stated, "Spatial prediction of debris flows in terms of susceptib ility mapping is the first and foremost requirement for disaster mitigation. In the present study, a comparative evaluation of machine learning and statistical approaches for debris flow susceptibility zonation (DFSZ) mapping has been attem pted using 10 causative thematic layers (slope, aspect, elevation, plan curvatur e, profile curvature, topographic wetness index, stream power index, geology, pr oximity to streams, normalized difference vegetation index) and a debris flow in ventory containing 85 debris flow locations." Our news journalists obtained a quote from the research from CSIR -Central Buil ding Research Institute, "The employed machine learning (ML) approaches include random forest (RF), na & iuml;ve Bayes (NB), and extreme gradient boosting (XGBoost) models whereas statistical models include the weight of evide nce (WoE) and index of entropy (IoE). The results indicated that in all 5 DFSZ m aps, about 21.20-47.98% of the area is very highly and highly susc eptible to debris flows. It is observed that the major debris flows as well as h igh susceptible zones are distributed along the river Alakananda and its tributa ries and at the vicinity of the NH-58. Among the statistical models, the DFSZ ma p prepared using the weight of evidence (WoE) model provides higher accuracy in terms of the success rate and the prediction rate compared to that prepared usin g the index of entropy model (IoE). Among the machine learning-based models, bot h the extreme gradient boosting (XGBoost) and random forest (RF) models give bet ter accuracy and are more efficient than the Na & iuml;ve Bayes (N B) model. It is also observed that the ML models perform better than the statist ical models for a part of Chamoli district, Uttarakhand state (India). The novel ty of the present study lies in the spatial prediction of one of the most destru ctive forms of mass movement (debris flow) in the Indian Himalayas using statist ical and ML models and establishing the superiority of the ML Random Forest & XGBoost model over other ML and statistical models for the present case. This st udy will help make decisions on the suitability of developmental activities and human settlement in the area under consideration."
RoorkeeIndiaAsiaCyborgsEmerging TechnologiesMachine LearningCSIR -Central Building Research Institute
NETL
NSTL
