Open AccessMeasuring the Impact of Statistical Techniques for Computation of Weighting Factors in Avalanche Forecasting Model
Author(s) -
Neha Ajit Kushe,
Ganesh Magar
Publication year - 2019
Publication title -
international journal of recent technology and engineering
Language(s) - English
Resource type - Journals
ISSN - 2277-3878
DOI - 10.35940/ijrte.d7500.118419
Subject(s) - weighting , rank correlation , correlation coefficient , snow , similarity (geometry) , spearman's rank correlation coefficient , statistics , pearson product moment correlation coefficient , work (physics) , computer science , econometrics , meteorology , environmental science , mathematics , artificial intelligence , engineering , geography , physics , acoustics , image (mathematics) , mechanical engineering
Avalanche forecasting is an important measure required for the safety of the people residing in hilly regions. Snow avalanches are caused due to the changes that occur in the snow and weather conditions. The prominent changes, that cause the variations which further culminate into an avalanche, can be given higher significance in the forecasting model by application of appropriate weights. These weights are decided based on the relation of each weather parameter to snow avalanche occurrence by the forecaster with the help of historical data. A method is proposed in the current work that can help in removing this subjectivity by using correlation coefficients. Present work explores the use of Pearson correlation coefficient, Spearman rank correlation coefficient and Kendall Tau correlation coefficient to obtain the weighting factors for each parameter used for avalanche forecasting. These parameters are further used in the cosine similarity based nearest neighbour model for avalanche forecasting. Bias and Peirce’s Skill Score are performance measures used to evaluate the outcome of the experimental work.