Author(s):
1. Cevher Özden:
Department of Computer Engineering, Akdeniz University,Antalya,Turkey
2. Nurgül Karadoğan:
Department of Agricultural Economics, Gaziosmanpasa University,Tokat,Turkey
Abstract:
Forecasting agricultural product yield is quite an important and elaborate task for agriculture sector. Previous information about future enables all parties included in agriculture sector to take necessary precautions to alleviate any possible damage. Wheat is possibly the most important food ingredient for many people in the world. It provides daily nutrition needs througho ut the world and is of strategical importance for the independence of many nations. The current study is carried out to analyze the applicability of various statistical, machine learning and deep learning methods on predicting wheat yield. For this purpose, weather and plant nutrient usage are used input variables and the wheat yield in the major producing provinces is considered as target output. The analysis results have demonstrated that all models are quite good at learning the relationship between the selected environment variables and wheat yield. However, models have achieved the highest accuracies in forecasting the wheat yield in Konya province. Furthermore, Random Forest ranked first in its prediction of wheat yield in Konya province. It is followed by CNN, Auto-Arima and LSTM methods.
Page(s):
429-435
Published:
Journal: Pakistan Journal of Agricultural Sciences, Volume: 61, Issue: 2, Year: 2024
Keywords:
Wheat
,
Random Forest
,
LSTM
,
CNN
,
statistical inference
,
Yield forecast
,
auto arima
,
seasonal autoregression
References:
References are not available for this document.
Citations
Citations are not available for this document.