Biblio

Mapping average annual precipitation in Serbia (1961–1990) by using machine learning models

TitleMapping average annual precipitation in Serbia (1961–1990) by using machine learning models
Publication TypeConference Paper
Year of Publication2014
AuthorsVišnjevac N, Kovačević M, Bajat B
Conference NameDailyMeteo.org/2014
Date Published06/2014
PublisherFaculty of Civil Engineering, University of Belgrade
Conference LocationBelgrade
Abstract

Precipitation data are measured at certain places, often quite distanced. Values between the places can be predicted using geostatistical tools like kriging, for which the interpolated values are modeled by a Gaussian process. In this paper we discuss an alternative method, i.e. using Machine learning techniques (Multilayer perceptron, linear regression and Support Vector Regression) to predict precipitation values. For the case study, average annual values of precipitation for a 30-year period in Serbia, three models were created based on attributes related to location data, spatial pattern, distance from nearby sea and data of nearest points (gauges). The best results were obtained by using Support Vector Regression and Linear Regression, indicating the linear nature of the problem. Finally, we have compared the output from kriging prediction and the output obtained by using Machine Learning techniques. 

Event Calendar

S M T W T F S
 
 
1
 
2
 
3
 
4
 
5
 
6
 
7
 
8
 
9
 
10
 
11
 
12
 
13
 
14
 
15
 
16
 
17
 
18
 
19
 
20
 
21
 
22
 
23
 
24
 
25
 
26
 
27
 
28
 
29
 
30
 
31
 
 
 

Meteo package

Recent comments

No comments available.