WSEAS Transactions on Computers
Print ISSN: 1109-2750, E-ISSN: 2224-2872
Volume 15, 2016
ISLIQ: Improved Supervised Learning in Quest to Nowcast Snow/No-Snow
Authors: , ,
Abstract: Nowcasting the presence of snow/no-snow is a major problem to most of the researchers, academicians, scientists and so on, as it would affects the lives of humans, animals and aquatic life, vegetation, tourism sector to a greater extent across the globe. Previously, many of the scientists, researchers, academicians and so on provided solutions but limited to the usage of Satellite imagery, Radar imagery and so on. We provided approaches related to the same, by making use of decision trees, but the main drawback with them is computational complexity during the evaluation of split points. With this focus, in this paper we are providing Improved Supervised Learning in Quest (ISLIQ) decision tree algorithm for the nowcasting of snow/no-snow, the evaluation of splitting criterion is based on interval range instead of computing whenever there is a change in the class label, which drastically reduces the number of split points. Further, we also evaluated the performance measures such as specificity, sensitivity, precision, dice, accuracy and error rate and compared the results with various decision tree algorithms.
Search Articles
Pages: 34-42
WSEAS Transactions on Computers, ISSN / E-ISSN: 1109-2750 / 2224-2872, Volume 15, 2016, Art. #4