COMPARISON AND ANALYSIS OF FORECASTING TECHNIQUES FOR ARRIVALS AND PRICES OF ONION IN HYDERABAD REGION
Loading...
Date
2023-03-20
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
PROFFESSOR JAYASHANKAR TELANGANA STATE AGRICULTURAL UNIVERSITY
Abstract
The edible bulb of the onion (Allium cepa), which is a member of the Liliaceae
or Amaryllidaceae family, is widely grown around the world. Additionally, it is a
biennial herbaceous vegetable crop and is one of the key vegetable crops that are eaten
year-round in every home. China, India, and the United States of America collectively
generate more than 50% of the world's total onion crop. Maharashtra, Madhya Pradesh,
Karnataka, Gujarat, and Rajasthan are the five states that produce the most onions in
India. A study of agricultural commodity price patterns has a special significance in
developing economies like India. When choosing when and where to sell the produce
from their farms, farmers are guided by price levels. Price swings and agriculture are
both well-known aspects of each. Considering the importance of onion in the economy
and as a vital agricultural commodity, it may be significant to study this crop in detail to
know the trend and variations in it. In the present study, the price and arrivals of Onions
are examined in three different markets such as Gudimalkapur, Bowenpally, and
Hyderabad markets undertaking three years of secondary daily data from 01-01-2019 to
31-12-2021 in Telangana State. The research is encapsulated with the primary
objectives of fitting linear and non-linear time series models and comparing them by
fitting different linear time series models like ARIMA, ARCH, and GARCH models
with the non-linear time series model ANN. The adequacy of the models is checked
through minimum RMSE, AIC, and MAPE values.
By comparing the models from linear and non-linear time series models, the best
model is selected based and forecasted values for January month. Comparing all three
markets the average arrivals and price was higher in the Hyderabad market compared to
Gudimalkapur and Bowenpally markets. The GARCH model among the linear time
series models was found to have the best fit for onion price, and the ARCH model for
onion arrival in Gudimalkapur market. Both in Bowenpally and Hyderabad markets for
onion price ARIMA (1, 1, 2) model was found to be the best among all linear time
series models fitted, and the ARCH model was best among onion arrivals with the least
model adequacy values. Comparing the actual and predicted values of the best-fitted
linear and non-linear time series models, shows that the ANN model predicted values
were more accurate at predicting future prices and arrivals of onions in all the markets
than the other models. This suggests that non-linear models performed better than linear
models