Wheat Yield Prediction for Turkey Using Statistical Machine Learning and Deep Learning Methods

Cevher Özden; Nurgül Karadoğan

doi:10.21162/PAKJAS/24.182

No Detail Found!

Wheat Yield Prediction for Turkey Using Statistical Machine Learning and Deep Learning Methods

Author(s)

Cevher Özden Department of Computer Engineering, Akdeniz University,Antalya,Turkey

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.

Publication Details

Page(s) 429-435

DOI 10.21162/PAKJAS/24.182

Published Journal: Pakistan Journal of Agricultural Sciences, Volume: 61, Issue: 2, Year: 2024

Request the full text from PASTIC

Keywords

Wheat, Random Forest, LSTM, CNN, statistical inference, Yield forecast, auto arima, seasonal autoregression.

Citations

Downloads

Views