STATISTICAL MODELS FOR FORCASTING ARRIVALS AND PRICES OF MAJOR VEGETABLES AT SELECTED MARKETS IN KURNOOL DISTRICT OF ANDHRA PRADESH
Loading...
Date
2024-03-06
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Acharya N G Ranga Agricultural University
Abstract
The Present study entitled “Statistical Models for Forecasting Arrivals and
Prices of Major Vegetables at Selected Markets in Kurnool District of Andhra
Pradesh” is mainly aimed at to study the trends, variability, association and to forecast
the arrivals and prices by best fitted model. Three major vegetable markets viz., C
Camp market, Adoni market and Nandyal market for two highly consumed vegetables
(tomato and onion) are considered for this study.
The secondary data of daily arrivals (qtls) and daily prices (Rs/Kg) were
collected from respective market committee for the period of three years two months
(January 2019 to February 2022).
The trend in the data was examined by observing trend lines.
In C Camp market, tomato arrivals and prices had fluctuating trend over days
during data period and peak arrivals (140 qtls) were noticed in the month of April-2020,
while lowest arrivals (30 qtls) were noticed in the month of July-2020, and peak price of
(74 Rs/Kg) was recorded in the month of November-2021, while least prices (6 Rs/Kg)
were noticed in the month of March-2019. Onion arrivals had showed a gradually
increasing trend and peak arrivals (80 qtls) were noticed in the month of January-2021,
but least arrivals (4 qtls) were noticed in October-2020 and prices had fluctuating trend
over the days in the data period and prices (70 Rs/Kg) were on hike in the month of
December-2019, and low prices (1 Rs/Kg) were recorded in the month of February 2021.
In Adoni market, tomato arrivals displayed a constant trend over data period but
peak arrivals (26 qtls) were recorded in the month of August-2020, low arrivals (14
qtls) were noticed in the month of February-2022 and peak prices (48 Rs/Kg) were
recorded in the month of March-2019 and prices had decreasing trend while least prices
(6 Rs/Kg) were noticed in April-2022, Onion arrivals had showed a gradually
decreasing trend and peak arrivals (18 qtls) were noticed in the month of March-2020,
least arrivals (1 qtl) were recorded in the month of January-2022 but onion prices had
xiii
fluctuating trend over the days in the data period and peak prices (80 Rs/Kg) were
noticed in the month of October-2020, and low prices (8 Rs/Kg) were recorded in the
month of February-2019.
In Nandyal market, tomato arrivals displayed a decreasing trend during the data
period but peak arrivals (32 qtls) were recorded in the month of September-2020, while
least arrivals (8 qtls) were noticed in the month of January-2019. Tomato prices were
showing fluctuating trend and peak prices (50 Rs/Kg) were noticed in the month of
June-2019, least prices (2 Rs/Kg) were recorded in the month of March-2020. Onion
arrivals had showed stable trend and peak arrivals (55 qtls) were noticed in the month of
January-2021, least arrivals (3 qtls) were noticed in the month of February-2019 but
onion prices had increasing trend in the data period and prices (56 Rs/Kg) were on hike
in the month of March-2020, least prices (4 Rs/Kg) were recorded in the month of
February-2019.
Coefficient of variation (CV) was calculated to understand the variation of
arrivals and prices of tomato and onion. For arrivals and prices, CV was almost non consistent with deviation over years, which indicated that the arrivals and price were
showing variation. But in case of Adoni market there was less-variation and consistency
in arrivals of tomato and onion.
To ascertain the relationship between arrivals and prices in respective market,
correlation coefficient was computed. There was a negative and significant correlation
between arrivals and prices of tomato and onion in all selected markets.
Auto Regressive Integrated Moving Average (ARIMA) models and Artificial
Neural Network (ANN) models were used for forecasting the arrivals and prices. At the
identification stage, one or more models are tentatively chosen and the most suitable
models are selected based on highest R2
and least RMSE and MAPE values. ANN
models outperformed ARIMA model in forecasting arrivals and prices in tomato and
onion at selected markets in Andhra Pradesh. The predicted values and actual values
were close to each other in most of the cases.