dc.contributor.author |
Mwanjele, S M |
|
dc.date.accessioned |
2017-05-18T12:49:34Z |
|
dc.date.available |
2017-05-18T12:49:34Z |
|
dc.date.issued |
2017-05-18 |
|
dc.identifier.isbn |
9966 923 28 |
|
dc.identifier.uri |
http://journals.jkuat.ac.ke/index.php/jscp/article/view/834 |
|
dc.identifier.uri |
http://hdl.handle.net/123456789/3085 |
|
dc.description.abstract |
The application of computer science has led to advancements in various sectors of economies
including agricultural production, manufacturing and marketing. Computer algorithms have been
used for prediction. There has been immense interest and research on meteorological prediction
aimed at addressing drought. This has been achieved through the development of various drought
indices. Some researchers have studied drought prediction by applying computer science solutions.
However, critical issues related to agricultural drought have not been well addressed. This study
looked at issues related to agricultural droughts, with the aim of developing an efficient and
intelligent agricultural drought prediction system. By using a case study approach and knowledge
discovery data mining process this study was preceded by literature review, followed by analysis of
daily 1978-2008 meteorological and annual 1976-2006 maize produce data in Voi Taita-Taveta
(Coast province, Kenya). The design and implementation of an agricultural drought prediction
system for meteorological data preprocessing, classification algorithms for training and testing as
well as prediction and post processing of predictions to various agricultural drought aspects is
accomplished. To overcome the problem of geographical differences the solution allows choice of
area latitude during the preprocessing. To come up with the agricultural drought meteorological
data relationships, the study was forced on the two different datasets. Meteorological data is on
daily basis while maize produce data is on annual basis. The datasets difference constraint was
overcome by performing analysis of metrological data on monthly, seasonal and yearly basis so as to
properly relate the two data sets. Further to overcome the limitation of data incompatibility the
analysis of each dataset was done independently. Literature review on drought occurrences verified
the results of associated maize produce and meteorological data analysis. Maize was used as a study
crop since it is the staple food and also most sensitive to agricultural drought compared to other
seasonal crops. The solution was evaluated by comparison of predicted to actual 2009 data and
Kenya Meteorological Department (KMD) records. The evaluation of our study results indicated
consistency with the KMD 2009 outlook. The report concludes that the application of classification
algorithms together with past meteorological data can lead to accurate predictions of future
agricultural drought. |
en_US |
dc.description.sponsorship |
JKUAT |
en_US |
dc.language.iso |
en |
en_US |
dc.publisher |
JKUAT |
en_US |
dc.relation.ispartofseries |
Proceedings of the 2012 JKUAT Scientific, Technological and Industrialization Conference;15-16th November |
|
dc.subject |
Data mining |
en_US |
dc.subject |
knowledge discovery |
en_US |
dc.subject |
classification algorithm |
en_US |
dc.subject |
intelligent prediction |
en_US |
dc.subject |
agricultural drought |
en_US |
dc.subject |
nearest neighbor classification |
en_US |
dc.subject |
Kenya |
en_US |
dc.subject |
JKUAT |
en_US |
dc.title |
INTELLIGENT PREDICTION OF AGRICULTURAL DROUGHT USING CLASSIFICATION ALGORITHMS |
en_US |
dc.type |
Article |
en_US |