AN APPLICATION OF HYBRID CLUSTERING AND NEURAL BASED PREDICTION MODELLING FOR DELINEATION OF MANAGEMENT ZONES
Journal Title: International Journal on Computer Science and Engineering - Year 2011, Vol 3, Issue 2
Abstract
Starting from descriptive data on crop yield and various other properties, the aim of this study is to reveal the trends on soil behaviour, such as crop yield. This study has been carried out by developing web application that uses a well known technique- Cluster Analysis. The cluster analysis revealed linkages between soil classes for the same field as well as between different fields, which can be partly assigned to crops rotation and determination of variable soil input rates. A hybrid clustering algorithm has been developed taking into account the traits of two clustering technologies: i) Hierarchical clustering, ii) K-means clustering. This hybrid clustering algorithm is applied to sensor- gathered data about soil and analysed, resulting in the formation of well delineated management zones based on various properties of soil, such as, ECa , crop yield, etc. One of the purposes of the study was to identify the main factors affecting the crop yield and the results obtained were validated with existing techniques. To accomplish this purpose, geo-referenced soil information has been examined. Also, based on this data, statistical method has been used to classify and characterize the soil behaviour. This is done using a prediction model, developed to predict the unknown behaviour of clusters based on the known behaviour of other clusters. In predictive modeling, data has been collected for the relevant predictors, a statistical model has been formulated, predictions were made and the model can be validated (or revised) as additional data becomes available. The model used in the web application has been formed taking into account neural network based minimum hamming distance criterion.
Authors and Affiliations
Babankumar S. Bansod , OP Pandey
Keywords
Related Articles
- EP ID EP108088
- DOI -
- Views 125
- Downloads 0
Generalisation of RSA Scheme using fundamental groups and ZKIP
We address the problem of computation involved in RSA algorithm namely exponentiation under modulo arithmetic and various mathematical and timing attacks in RSA. The computation is made easy and quick by assigning elemen...
Ontology driven Pre and Post Ranking based Information Retrieval in Web Search Engines
With the tremendous growth of World Wide Web, it has become necessary to organize the information in such a way that it will make easier for the end users to find the information they want efficiently and accurately. Thi...
Mobile Software Testing – Automated Test Case Design Strategies
Mobile devices are poised to challenge PCs as the application platform of choice, with 500 million mobile internet devices expected to ship in 2012 compared to 150 million PCs. The convergence of all digital devices into...
Backpropagation Learning Algorithms for Email Classification.
Today email has become one the fastest and most effective form of communication. The popularity of this mode of transmitting goods, information and services has motivated spammers to perfect their technical skills to foo...
AN IMPROVED LOW COMPLEX SPATIALLY SCALABLE ACC-DCT BASED VIDEO COMPRESSION METHOD
In this paper, we propose a low complex Scalable ACC-DCT based video compression approach which tends to hard exploit the pertinent temporal redundancy in the video frames to improve compression efficiency with less proc...