.

ISSN 2063-5346
For urgent queries please contact : +918130348310

CONCEPTUAL CLIMATIC CONDITION PREDICTION FOR CROP-BASED CULTIVATION USING SVM ALGORITHM COMPARED WITH LINEAR REGRESSION.

Main Article Content

Yaragadla Sivaram, Aishwarya B
» doi: 10.31838/ecb/2023.12.sa1.415

Abstract

Aim: In this study, the Support Vector Machine (SVM) algorithm is used to predict climatic parameters of the crops and its performance is tested by comparing it with the Linear Regression for the crop-based cultivation on climatic parameters for the yield prediction. Materials and Methods: The Crop-based on Climatic Parameters consists of 3101 of different crops and climatic parameters used for training 3000 (80%) and testing 101 (20%) the predictive model in python and the statistical analysis is done using SPSS software. The sample size is estimated using G-power analysis to be 3101 records in each group with 80% of power and a 0.05 Error rate. SVM algorithm is used and compared with Linear Regression algorithm. Results: The Support Vector Machine (SVM) algorithm’s predictive model shows a higher accuracy of 83.70% than the Linear Regression based model with an accuracy of 88.95% and with a significance value 0.012 (p<0.05). Conclusion: Within the limits of study confirms that the Linear Regression based model provides more promising results in the Crop-based cultivation on Climatic parameters than the Support Vector Machine (SVM) based model.

Article Details