A Deep Learning Model for Soybean Yield Prediction

EasyChair Preprint 10612

Abstract

Over the past two decades in India there is a climate has a significant impact on agricultural crops. This project will enable farmers in determining the yield of their crop prior to cultivating the agricultural land, allowing them to make better decisions. In this work soybean yield data is collected that contains around 6 features of soil data, 10 features of weather data, yield performance and management data. In this research a deep learning framework, CNN is proposed for efficient soybean yield prediction. We tested our proposed model against a variety of models, including the ANN, LSTM. The MSE of our proposed CNN model was 20.74, the RMSE was 4.55, the MAE was 3.34, and the R2 was 0.83. This comparison results demonstrated that our model CNN outperformed all other methods and is more effective in predicting soybean.

Keyphrases: ANN, Agriculture, CNN, Farmer, LSTM, Mean Absolute Error, Root Mean Square Error

@booklet{EasyChair:10612,
  author    = {Sindhu Yadulla and Shashi Vardhan Yasa and Nithin Pothu},
  title     = {A Deep Learning Model for Soybean Yield Prediction},
  howpublished = {EasyChair Preprint 10612},
  year      = {EasyChair, 2023}}