Niketa Gandhi

Rice crop yield prediction in India using support vector machines

Food production in India is largely dependent on cereal crops including rice, wheat and various pulses. The sustainability and productivity of rice growing areas is dependent on suitable climatic conditions. Variability in seasonal climate conditions can have detrimental effect, with incidents of drought reducing production. Developing better techniques to predict crop productivity in different climatic conditions can assist farmer and other stakeholders in better decision making in terms of agronomy and crop choice. Machine learning techniques can be used to improve prediction of crop yield under different climatic scenarios. This paper presents the review on use of such machine learning technique for Indian rice cropping areas. This paper discusses the experimental results obtained by applying SMO classifier using the WEKA tool on the dataset of 27 districts of Maharashtra state, India. The dataset considered for the rice crop yield prediction was sourced from publicly available Indian Government records. The parameters considered for the study were precipitation, minimum temperature, average temperature, maximum temperature and reference crop evapotranspiration, area, production and yield for the Kharif season (June to November) for the years 1998 to 2002. For the present study the mean absolute error (MAE), root mean squared error (RMSE), relative absolute error (RAE) and root relative squared error (RRSE) were calculated. The experimental results showed that the performance of other techniques on the same dataset was much better compared to SMO.

Effective ICTs in agricultural value chains to improve food security: An international perspective

This paper examines the grains value chain in agriculture, and identifies the importance in developing strategies which could better secure food production. The study highlights examples of successful integration of ICTs in agricultural supply and value chains. The development of strategies to integrate these ICTs into the supply chain will be proposed. It will be argued that the use of high powered computing for data mining and other technologies such as sensor networks, mobile communications, and GPS technologies can revolutionize the efficiency of these supply chains and therefore improve the food security. The study carried out a situational analysis of agricultural resources using standard internet search engines and applying data mining techniques in order to demonstrate how such technologies can be used to show difference in value chains across different situations. An assessment of the study found that the results from the grain-industry dataset support the similar supply chain grouping reported for other research studies. These groupings reflect the more-developed food-industry supply chains and may not capture all the interactions in less-developed supply chains. For example, when several of the food production processes are carried out by one food-producer, the activities will be more difficult to identify.

Use of Information and Communication Technology (ICT) Tools by Rural Farmers in Ratnagiri District of Maharastra, India

This paper discusses the impact of information and communication technology (ICT) on the access for rural farmers from the Ratnagiri district to agricultural information. A study was undertaken in which more than one hundred randomly selected farmers completed a structured questionnaire to gather information at household level of the use of ICT. Interviews were also conducted with key stakeholders, service providers and government officials. Findings from the study indicated that farmers were most interested in obtaining market price information. Examination of the relationship between use of ICT tools and co factors such as age, qualifications and income indicated that only income was a determining factor of using ICT tools.

PredictingRice crop yield using Bayesian networks

Rice crop production plays a vital role in food security of India, contributing more than 40% to overall crop production. High crop production is dependent on suitable climatic conditions. Detrimental seasonal climate conditions such as low rainfall or temperature extremes can dramatically reduce crop yield. Developing better techniques to predict crop productivity in different climatic conditions can assist farmer and other stakeholders in important decision making in terms of agronomy and crop choice. This paper reports on the use of Bayesian Networks to predict rice crop yield for Maharashtra state, India. For this study, 27 districts of Maharashtra were selected on the basis of available data from publicly available Indian Government records with various climate and crop parameters selected. The parameters selected for the study were precipitation, minimum temperature, average temperature, maximum temperature, reference crop evapotranspiration, area, production and yield for the Kharif season (June to November) for the years 1998 to 2002. The dataset was processed using the WEKA tool. The classifiers used in the study were BayesNet and NaiveBayes. The experimental results showed that the performance of BayesNet was much better compared with NaiveBayes for the dataset.

Rice crop yield prediction using artificial neural networks

Rice crop production contributes to the food security of India, more than 40% to overall crop production. Its production is reliant on favorable climatic conditions. Variability from season to season is detrimental to the farmers income and livelihoods. Improving the ability of farmers to predict crop productivity in under different climatic scenarios, can assist farmers and other stakeholders in making important decisions in terms of agronomy and crop choice. This study aimed to use neural networks to predict rice production yield and investigate the factors affecting the rice crop yield for various districts of Maharashtra state in India. Data were sourced from publicly available Indian Governments records for 27 districts of Maharashtra state, India. The parameters considered for the present study were precipitation, minimum temperature, average temperature, maximum temperature and reference crop evapotranspiration, area, production and yield for the Kharif season (June to November) for the years 1998 to 2002. The dataset was processed using WEKA tool. A Multilayer Perceptron Neural Network was developed. Cross validation method was used to validate the data. The results showed the accuracy of 97.5% with a sensitivity of 96.3 and specificity of 98.1. Further, mean absolute error, root mean squared error, relative absolute error and root relative squared error were calculated for the present study. The study dataset was also executed using Knowledge Flow of the WEKA tool. The performance of the classifier is visually summarized using ROC curve.

A review of the application of data mining techniques for decision making in agriculture

This paper provides a review of research on the application of data mining techniques for decision making in agriculture. The paper reports the application of a number of data mining techniques including artificial neural networks, Bayesian networks and support vector machines. The review has outlined a number of promising techniques that have been used to understand the relationships of various climate and other factors on crop production. This review proposes that further investigations are needed to understand how these techniques can be used with complex agricultural datasets for crop yield prediction integrating seasonal and spatial factors by using GIS technologies.

Proposed decision support system (DSS) for Indian rice crop yield prediction

Rice crop production provides more than 40% to overall crop production in India and is essential in ensuring food security. Its production is reliant on favorable climatic conditions. Improving the ability of farmers to predict crop productivity under different climatic scenarios, can assist farmers and other stakeholders in making important decisions in terms of agronomy and crop choice. This paper proposes a decision support system prototype for rice crop yield prediction for Maharashtra state, India. A Graphical User Interface (GUI) has been created in Java using NetBeans tool and Microsoft Office Access database for the ease of farmers and decision makers. The interface allows for the selection of the range of precipitation, minimum temperature, average temperature, maximum temperature and reference crop evapotranspiration and predicts the expected class of yield viz., low, moderate or high. The ranges of the parameters were calculated by using historic data from the study area. The classes for the yield were defined as low with 0.15 to 0.60 tonnes/hectare; moderate with 0.61 to 1.10 tonnes/hectare and high with 1.11 to 3.16 tonnes/hectare. The proposed prototype could be used for a bigger dataset and wider study area to predict the crop yield. This will provide a guide to the farmer to assist in decision making on potential crop yield for particular climatic scenario.

Rice crop yield forecasting of tropical wet and dry climatic zone of India using data mining techniques

Data mining is the process of identifying the hidden patterns from large and complex data. It may provide crucial role in decision making for complex agricultural problems. Data visualisation is also equally important to understand the general trends of the effect of various factors influencing the crop yield. The present study examines the application of data visualisation techniques to find correlations between the climatic factors and rice crop yield. The study also applies data mining techniques to extract the knowledge from the historical agriculture data set to predict rice crop yield for Kharif season of Tropical Wet and Dry climatic zone of India. The data set has been visualised in Microsoft Office Excel using scatter plots. The classification algorithms have been executed in the free and open source data mining tool WEKA. The experimental results provided include sensitivity, specificity, accuracy, F1 score, Mathews correlation coefficient, mean absolute error, root mean squared error, relative absolute error and root relative squared error. General trends in the data visualisation show that decrease in precipitation in the selected climatic zone increases the rice crop yield and increase in minimum, average or maximum temperature for the season increases the rice crop yield. For the current data set experimental results show that J48 and LADTree achieved the highest accuracy, sensitivity and specificity. Classification performed by LWL classifier displayed the lowest accuracy, sensitivity and specificity results.

Assessing impact of seasonal rainfall on rice crop yield of Rajasthan, India using association rule mining

Developing countries which are highly dependent on agriculture have shown growing concern that climate variability will further impact on food security. It is important to have a deeper understanding of the impact of this climate change on crop production and food security. This paper assesses the impact of distributed seasonal rainfall on rice crop yield of Rajasthan state, India through data visualisation and application of association rule mining techniques. The dataset considered for the present study was of twenty nine districts of Rajasthan state for forty three years from 1960 to 2002 depending on the data availability. Three divisions were made for the rainfall in Kharif season from June to November. Beginning of the season was considered as June and July, Middle of the season as August and September and End of the season as October and November. The effect of variation in the rainfall at the beginning, middle and end of season on the rice crop yield was investigated and some interesting results are reported.

Automatic identification of Malaria using image processing and artificial neural network

Malaria is a mosquito-borne infectious disease, which is diagnosed by visual microscopic assessment of Giemsa stained blood smears. Manual detection of malaria is very time consuming and inefficient. The automation of the detection of malarial cells would be very beneficial in the treatment of patients. This paper investigates the possibility of developing automatic malarial diagnosis process through the development of a Graphical User Interface (GUI) based detection system. The detection system carries out segmentation of red blood cells (RBC) and creates a database of these RBC sample images. The GUI based system extracts features from smear image which were used to execute a segmentation method for a particular blood smear image. The segmentation technique proposed in this paper is based on the processing of a threshold binary image. Watershed threshold transformation was used as a principal method to separate cell compounds. The approach described in this study was found to give satisfactory results for smear images with various qualitative characteristics. Some problems were noted with the segmentation process with some smear images showing over or under segmentation of cells. The paper also describes the feature extraction technique that was used to determine the important features from the RBC smear images. These features were used to differentiate between malaria infected and normal red blood cells. A set of features were proposed based on shape, intensity, contrast and texture. These features were used for input to a neural network for identification. The results from the study concluded that some features could be successfully used for the malaria detection.