首页|Findings from University of Kentucky in Machine Learning Reported(Coupled Lands lide Analyses Through Dynamic Susceptibility andForecastable Hazard Analysis)
Findings from University of Kentucky in Machine Learning Reported(Coupled Lands lide Analyses Through Dynamic Susceptibility andForecastable Hazard Analysis)
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
By a News Reporter-Staff News Editor at Robotics & Machine Learning DailyNews Daily News – Current study results on Machine Learn ing have been published. According to newsreporting out of Lexington, Kentucky, by NewsRx editors, research stated, “Landslides, specifically thosetriggered t hrough an increase of soil moisture either during or after a rainfall event, pos e severe threatsto surrounding infrastructure. Herein, the term ‘landslide’ ref ers primarily to translational movements ofshallow colluvial soil upon a hillsl ope.”Our news journalists obtained a quote from the research from the University of K entucky, “Theselandslides are assumed to adhere to infinite slope approximation s. Potential landslide occurrences aremonitored through identification of areas susceptible to occurrence, through susceptibility analyses, or areaslikely to experience a landslide at a given time, through hazard analyses. Traditional lan dslide susceptibilitysystems are created as a function of static geomorphologic data. This is to say that, while spatially differing,susceptibility via this s ystem does not change with time. Landslide hazard analyses consider dynamic data, such as that of precipitation, and provide warnings of when landslide occurren ces are likely. However,these hazard analysis systems typically only provide wa rnings in near real time (i.e., over the next fewdays). Therefore, dynamic susc eptibility (susceptibility that is seen to change with time rather than remains tatic) as well as the ability to forecast landslide hazard analyses beyond real time is desired. The studyherein presents a novel workflow for the creation of dynamic landslide susceptibility and forecastable hazardanalyses over a domain within Eastern Kentucky. Dynamic susceptibility was developed through inclusionof static geomorphic parameters and dynamic vegetation levels over sites of inte rest. These susceptibilitydata were used in the development of a logistic regre ssion classification machine learning approach whichyielded susceptibility clas sifications with an accuracy of 89%. Forecastable hazard analyses w ere developedas a function of forecasted soil moisture, assumed to be a control ling factor in landslide occurrence, overa site. Forecasting of soil moisture w as conducted through development of a Long Short-Term Memory(LSTM) forecasting machine learning system. Forecasts of soil moisture were then assimilated into a ninfinite slope stability equation to provide forecasts of hazard analyses. The se forecasted hazard analyseswere investigated over known landslides with satis factory results obtained.”
LexingtonKentuckyUnited StatesNort h and Central AmericaCyborgsEmerging TechnologiesMachine LearningUnivers ity of Kentucky
NETL
NSTL
