Deep Learning Based Prediction on Greenhouse Crop Yield Combined TCN and RNN.

Area of Science:

• Agricultural Science
• Computer Science
• Machine Learning

Background:

• Greenhouse cultivation relies on controlled environmental parameters for optimal crop yield.
• Accurate crop yield prediction is essential for effective greenhouse farming planning and financial decision-making.

Purpose of the Study:

• To develop an advanced technique for precise greenhouse crop yield prediction.
• To enhance the accuracy of yield forecasting by integrating novel deep learning models.

Main Methods:

• A hybrid deep learning model combining Temporal Convolutional Network (TCN) and Recurrent Neural Network (RNN) was developed.
• The proposed model was evaluated on multiple real-world greenhouse datasets for tomato cultivation.
• Performance was assessed by comparing predicted yields against actual yields using root mean square error (RMSE).

Main Results:

• The proposed TCN-RNN hybrid model demonstrated superior accuracy in crop yield prediction compared to traditional machine learning and other deep neural networks.
• Statistical analysis confirmed the enhanced predictive performance of the novel approach.
• Historical yield data was identified as the most significant factor influencing future yield prediction accuracy.

Conclusions:

• The developed TCN-RNN model offers a more accurate method for greenhouse crop yield prediction.
• Accurate yield forecasting using this technique can significantly benefit greenhouse farming management and decision-making.
• Future research should leverage historical yield data for further optimization of prediction models.