International Journal of Knowledge Based Computer System

Sadakshya Sharma ¹, Jeevan Singh Bhasin ¹, Ashish ¹, Haneet Kour ², Karun Handa ³

Affiliations
1 B.Tech. Final Year Student, Department of Computer Science and Engineering, PIET, Haryana, IN
2 Ph.D. Research Scholar, Department of Computer Science and IT, University of Jammu, J&K, IN
3 Assistant Professor, Department of Computer Science and Engineering, PIET, Haryana, IN

Abstract

Machine learning techniques play an important role in solving real world problems. These techniques are also found to be successful in the field of Agriculture for crop yield prediction, leaf disease detection, fruit disease detection, vegetable quality assessment, etc. In this paper, the authors performed comparative analysis of various supervised machine learning techniques for crop yield prediction from soil parameters. Five supervised machine learning techniques such as K-Nearest Neighbor (KNN), Support Vector Machine (SVM), Naïve Bayes (NB), Decision Tree (DT) and Random Forest (RF) have been taken for the experimental analysis. The experiments have been carried out for the prediction of the three most commonly grown crops in India: Rice, Wheat and Mustard. The performance of each technique for every crop taken in this study, has been evaluated on the basis of four metrics i.e. accuracy, recall, precision and f-score. The experimental results revealed that decision tree and random forest performed better than all the other supervised machine learning techniques taken in this study, for the prediction of each crop.

Keywords

DT, KNN, Mustard, RF, Rice, SVM, Wheat

Full Text

     

Yukti Gupta ¹, Haneet Kour ², Jatinder Manhas ³, Vinod Sharma ⁴

Affiliations
1 M.Tech. Student, Department of Computer Science and IT, University of Jammu, Jammu and Kashmir, IN
2 PhD Research Scholar, Department of Computer Science and IT, University of Jammu, Jammu and Kashmir, IN
3 Sr. Assistant Professor, Department of Computer Science and IT, Bhaderwah Campus, University of Jammu, Jammu and Kashmir, IN
4 Professor, Department of Computer Science and IT, University of Jammu, Jammu and Kashmir, IN

Abstract

Machine learning (ML) has played a significant role in pattern recognition including fruits and vegetables classification. In this paper, comparative analysis of various ML techniques have been carried out for the identification of Spices. For the current work, ML techniques namely Naïve Bayes (NB), Decision Tree (DT), K-Nearest Neighbor (KNN), Random Forest (RF) and Support Vector Machine (SVM) have been undertaken. The main aim of the current study is to find out the most appropriate ML approach for Spices recognition. The experimental study has been performed on primary dataset of Spices. This dataset consists of 1000 images of five different Spices including clove, green cardamom, cinnamon, black pepper and curry leaf. The performance of the ML techniques have been analyzed on the basis of four parameters i.e. accuracy, precision, recall and f1-score. Out of five implemented ML models, best performance has been predicted by SVM approach with accuracy of 94.5%, precision of 95%, and recall of 94% with f1-score of 0.95..

Keywords

Decision Tree, K-Nearest Neighbor, Machine Learning, Spices Recognition, Support Vector Machine

Full Text