Forecasting the Crop Yield Production in Trichy District Using Fuzzy C-Means Algorithm and Multilayer Perceptron (MLP)

Forecasting the Crop Yield Production in Trichy District Using Fuzzy C-Means Algorithm and Multilayer Perceptron (MLP)

Geetha M. C. S. (Kumaraguru College of Technology, India) and Elizabeth Shanthi I. (Avinashilingam Institution for Home Science and Higher Education for Women, India)
Copyright: © 2020 |Pages: 16
DOI: 10.4018/IJKSS.2020070105
OnDemand PDF Download:
Available
$37.50
No Current Special Offers
TOTAL SAVINGS: $37.50

Abstract

The agricultural stock depends upon several factors like biological, seasonal, and economic determinants. The growers sustain a vital loss if they are not capable of predicting the variations in these circumstances. The uncertainty on crop yield can be predicted in a logical and mathematical way. The forecast is made based on the previous archives of yield data secured from that area. Data mining is one such procedure practised to predict the crop yield. The systems examine the data, and on mining, several patterns based on numerous parameters predict the return. This article directs on crop yield forecast in Trichy district by adopting data mining techniques for rule formation on classifying the training data and implementing prediction for test data. The suggested method employs fuzzy C means algorithm for clustering and multilayer perceptron design for prediction. The results of accuracy and execution time of the proposed system correlated with the regression algorithm of prediction.
Article Preview
Top

1. Introduction

Agriculture offers 65 to 70% of the Indian economy, and therefore, it is the backbone of India. From 60.3% of agricultural land, it renders 17% of GDP and 10% of entire commodity. Agriculture support 60% of people's job. It boosts the foreign market, but farmers suffer from not perceiving required yield due to numerous causes. The computer procedures are employed to subdue crop production obstacles in predicting the yield and risks in advancement. The approaches are applied to the massive collection of agricultural raw data from which beneficial information and patterns have been obtained (Papageorgiou et al., 2010).

More than half types of crops depend on monsoon and yield according to that. So farmers are interested in knowing the predictions. As for the growing population in India, agricultural growth is significant to meet the demand, the research on applying computer techniques is essential. The crop yield is very much influenced by factors like atmosphere temperature, rainfall and geographical topology and many others. A very exhaustive analysis is made on extensive data of the agriculture sector. To protect the crops from the natural disaster, the farmer is interested in analysing the seasonal changes. Then the prediction about the yield from the past nearby field data can be made and which can be helped using data mining methods.

Data mining is practised comprehensivly in the agricultural field for obtaining intelligence from raw data, and hidden patterns oriented to various parameters are derived. The necessary steps to using data mining in knowledge extraction are (Cunningham & Holmes, 1999):

  • 1.

    Data collection;

  • 2.

    Data pre-processing with cleaning and formatting data;

  • 3.

    Processing training and testing data with suitable machine learning algorithm for classification and knowledge extraction.

The data mentioned above mining necessary steps are depicted in the Figure 1.

Figure 1.

Data mining process in knowledge extraction

IJKSS.2020070105.f01

The classification of crop data based on various factors is done using supervised learning methods. Some of the machine learning algorithms is supervised in which the membership is predicted from test cases to label the remaining training cases. The unsupervised method is used for clustering predicts which new cases with features. And it is also used in exploratory data analysis when no preconception is available (Ramesh & Vardhan, 2013). Data mining algorithms such as K- Nearest neighbour, Kmeans, neural networks and SVM are practised in agricultural data for data analysis and prediction study.

1.1. Motivation

Agriculture is a prime source for the survival of the expanding Indian community. Diverse seasonal, commercial and biological patterns affect the crop production, but unpredictable changes in these patterns lead to a significant loss to farmers. These risks can be reduced when suitable approaches are employed on data related to soil type, temperature, atmospheric pressure, humidity and crop type. The crop yield depends on multiple different factors such as climate changes, soil type etc. Farmers are interested in knowing the crop yield beforehand. Traditionally, this process was dependent on the experiences of farmers, and it used to be limited only for a particular region.

The unpredictability and delay in the present production estimations are posing severe problems for planners to take timely import-export decisions. Therefore, consistent prediction methods are needed to help planners and policymakers make strategic decisions to safeguard national interest. Transfer of technology is one of the essential areas to be addressed. Most of the work so far carried out concentrates on different dimensions. A very few works have been carried out using the promising field of machine learning.

In the section 1, we discussed the applications of data mining in agriculture. In section 2, we described the related work. Section 3 provides materials and methods in this field. Section 4 contains the proposed methodology. Section 5 contains the results and discussion. And finally, the conclusion was given in section 6.

Complete Article List

Search this Journal:
Reset
Open Access Articles: Forthcoming
Volume 13: 4 Issues (2022): Forthcoming, Available for Pre-Order
Volume 12: 4 Issues (2021): 2 Released, 2 Forthcoming
Volume 11: 4 Issues (2020)
Volume 10: 4 Issues (2019)
Volume 9: 4 Issues (2018)
Volume 8: 4 Issues (2017)
Volume 7: 4 Issues (2016)
Volume 6: 4 Issues (2015)
Volume 5: 4 Issues (2014)
Volume 4: 4 Issues (2013)
Volume 3: 4 Issues (2012)
Volume 2: 4 Issues (2011)
Volume 1: 4 Issues (2010)
View Complete Journal Contents Listing