DSpace Repository

Evaluation of data mining techniques to forecast floods in the Bagmati river basin, Nepal.

Show simple item record

dc.contributor.advisor Guha, Prof. Sumanta
dc.contributor.author Maskay, Shristi
dc.contributor.other Esichaikul, Dr. Vatcharaporn
dc.contributor.other Shrestha, Dr. Sangam
dc.date.accessioned 2015-05-11T01:43:45Z
dc.date.available 2015-05-11T01:43:45Z
dc.date.issued 2015-05-10
dc.identifier.other AIT
dc.identifier.uri http://www.cs.ait.ac.th/xmlui/handle/123456789/739
dc.description 67 p en_US
dc.description.abstract Nepal witnesses a heavy loss of life and property, every year due to massive floods, landslides and avalanches. This has a direct impact on the overall development of the country. Therefore, to minimize the casualties and damage it is important to focus on implementing techniques for the prediction of extreme environmental and climatic conditions. Thus, this study has been developed with the aim to try to predict such hazards. The study area taken is Bagmati river basin located in Central Nepal and the main objective is to predict flooding. The main objective is the use of different data mining tools and techniques; in particular, multiple linear regression, support vector regression, and artificial neural network with feedforward back propagation to find the correlation between the rainfall and the discharge and to predict the flood in the Bagmati River Basin. Different data mining tools were applied to the discharge and precipitation data to predict the discharge at the outlet station. A trial-and-error process was used to select model parameters and to select the best model for prediction. Evaluation methods, in particular, root mean square error, coefficient of determination and Nash-Sutcliffe model efficiency coefficient analysis were used to determine the accuracy of the models. The result obtained from the evaluation method and the various predicted hydrograph of each model proved ANN model to be better in prediction of discharge though ANN model is considered to be one of the most time consuming models for prediction. The training of model even the testing highly depends on precision of the sample data selected. The outcome depends on the correlation of the attributes being used in the dataset and the analysis done on basis of historical records of rainfall and discharge of the various selected hydrological meteorological station of Bagmati river basin. en_US
dc.description.sponsorship AIT Fellowship en_US
dc.language.iso en en_US
dc.publisher AIT en_US
dc.subject Data Mining en_US
dc.subject Flood prediction en_US
dc.subject flood forecasting en_US
dc.subject data mining tools and techniques en_US
dc.subject SVM en_US
dc.subject ANN en_US
dc.subject Regression en_US
dc.title Evaluation of data mining techniques to forecast floods in the Bagmati river basin, Nepal. en_US
dc.type Research report en_US


Files in this item

This item appears in the following Collection(s)

Show simple item record

Search DSpace


Advanced Search

Browse

My Account