Treffer: AN INTELLIGENT DEEP LEARNING FRAMEWORK FOR MANGO YIELD PREDICTION IN PRECISION AGRICULTURE.

Predicting crop yields, particularly for high-value horticultural crops such as mangos, is critical for optimizing horticultural practices, particularly given the complexity of high-dimensional, nonlinear interactions between agronomic and environmental factors. This research develops an intelligent predictive framework using Artificial Neural Networks (ANNs) and deep learning algorithms to address the challenges of data complexity and temporal variability. The ANN architecture consists of an input layer with eight neurons representing critical agronomic inputs, a hidden layer of ten neurons with a ReLU activation function for non-linearity and a single linear output neuron for yield prediction. The problem is mathematically framed as minimizing the loss function to capture intricate temporal and nonlinear relationships affecting yield, and the model is trained using the back-propagation algorithm. The framework compares feedforward neural networks, recurrent neural networks (RNN), and long short-term memory (LSTM) networks to approximate the nonlinear mappings between inputs and yield. The framework was trained and evaluated using an 80:20 train-test split, achieving a prediction accuracy of 78%, with LSTM outperforming Feedforward Neural Networks (FNNs) and Recurrent Neural Networks (RNNs). The evaluation metrics used include Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and R2 score, ensuring the reliability of the predictions. The study introduces a desktop GUI user-friendly application developed in a Python (spyder) environment, libraries for yield predictions, and an android mobile app for management practices for farmers and stakeholders. These tools bridge advanced AI methodologies and practical agricultural needs, promoting data-driven decision-making in sustainable farming practices. [ABSTRACT FROM AUTHOR]

Copyright of International Journal of Agricultural & Statistical Sciences is the property of Kirti Shree and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)