Rathod, S., Mishra, G. C. (2018). Statistical Models for Forecasting Mango and Banana Yield of Karnataka, India. , 20(4), 803-816.
S. Rathod; G. C. Mishra. "Statistical Models for Forecasting Mango and Banana Yield of Karnataka, India". , 20, 4, 2018, 803-816.
Rathod, S., Mishra, G. C. (2018). 'Statistical Models for Forecasting Mango and Banana Yield of Karnataka, India', , 20(4), pp. 803-816.
Rathod, S., Mishra, G. C. Statistical Models for Forecasting Mango and Banana Yield of Karnataka, India. , 2018; 20(4): 803-816.

Statistical Models for Forecasting Mango and Banana Yield of Karnataka, India

Journal of Agricultural Science and Technology
Article 13, Volume 20, Issue 4, 2018, Pages 803-816    PDF (495.37 K)
Authors
S. Rathod* 1; G. C. Mishra2
1ICAR-Indian Agricultural Statistics Research Institute, New Delhi, India-11002.
2Institute of Agricultural Sciences, Banaras Hindu University, Varanasi, UP, India-221005.
Abstract
Horticulture sector plays a prominent role in economic growth for most of the developing countries. India is the largest producer of fruits and vegetables in the world next only to China. Among the horticultural crops, fruit crops are cultivated in majority of the area. Fruit crops play a significant role in the economic development, nutritional security, employment generation, and overall growth of a country. Among fruit crops, mango and banana are largest producing fruits of India. Generally, Karnataka is called as the horticultural state of India. In Karnataka, mango and banana are highest producing fruit crops. With these prospective, yield of mango and banana of Karnataka have been chosen as study variables. Forecasting is a primary aspect of developing economy so that proper planning can be undertaken for sustainable growth of the country. In this study, classes of linear and nonlinear, parametric and non-parametric statistical models have been employed to forecast yield of mango and banana of Karnataka. The major drawback of linear models is the presumed linear form of the model. In most of the cases, the time series are not purely linear or nonlinear as they contain both linear and nonlinear components. To overcome this problem a hybrid model has been proposed which consists of linear and nonlinear models. The hybrid model with the combination of Autoregressive Integrated Moving Average (ARIMA) and Support Vector Regression model performed better in both model building as well as in model validation as compared to other models.
Keywords
ARIMA; Hybrid models; NLSVR; Regression model; Time Series; TDNN
References
Statistics
Article View: 137
PDF Download: 110
Statistics
Number of Journals45
Number of Issues2,163
Number of Articles24,582
Article View20,501,470
PDF Download16,311,963