Machine Learning Algorithms for Oilseed Disease Diagnosis

5 Pages Posted: 17 Apr 2019

See all articles by Archana Thakur

Archana Thakur

Institute of Engineering and Technology Devi Ahilya University

Ramesh Thakur

Institute of Engineering and Technology Devi Ahilya University

Date Written: April 15, 2019

Abstract

Classification is a process of observing the characteristics of a new object and consigning it to a known class. The classification problem is distinguished by known classes and a training set of containing pre-categorized samples. The work presents machine learning classification algorithms viz. simple logistic, decision tree, random forest and multilayer perceptron for exact identification of oilseed diseases. Experiments are conducted using 10-fold cross validation. The results recommend that aforementioned classification algorithms diagnose the oilseed diseases at a significant accuracy level. Simple logistic and multilayer perceptron perform better than other algorithms and yield 96.33% and 95.90% accuracy. Random forest and decision tree result in providing fair results in less time. The presented algorithms facilitated in characterization of oilseed diseases using important variables and permit enhanced interpretability.

Keywords: Machine learning, Simple logistic, Decision tree, Multilayer perceptron, Oilseed diseases

Suggested Citation

Thakur, Archana and Thakur, Ramesh, Machine Learning Algorithms for Oilseed Disease Diagnosis (April 15, 2019). Proceedings of Recent Advances in Interdisciplinary Trends in Engineering & Applications (RAITEA) 2019. Available at SSRN: https://ssrn.com/abstract=3372216 or http://dx.doi.org/10.2139/ssrn.3372216

Archana Thakur (Contact Author)

Institute of Engineering and Technology Devi Ahilya University ( email )

Indore, 452001
India

Ramesh Thakur

Institute of Engineering and Technology Devi Ahilya University ( email )

Indore, 452001
India

Register to save articles to
your library

Register

Paper statistics

Downloads
7
Abstract Views
66
PlumX Metrics