Leaf area estimation in Coffea canephora genotypes by neural networks and multiple regression/Estimativa da area foliar de genotipos de Coffea canephora por meio de redes neurais e regressao multipla

• Published in: Revista brasileira de engenharia agrícola e ambiental 2024-09, Vol.28 (9), p.1
• Main Authors: da Vitoria, Edney L, Junior, Andre O. Nardotto, Ribeiro, Luis F.O, Dubberstein, Danielly, Partelli, Fabio L
• Format: Article
• Language: English
• ISSN: 1415-4366
• DOI: 10.1590/1807-1929/agriambi.v28n9e279246

Leaf area data from coffee plants are important for studies and analyses of grain yield, physiology adaptation to environmental conditions, and cultural management. The objective of this study was to predict leaf area of coffee plants using artificial neural networks and compare the efficiency of this methodology with multiple regression models. Forty-three genotypes of similar reproduction and age were evaluated, testing 14 types of multiple regression equations from combinations of leaf length and width. The backpropagation algorithm was used to develop multilayer perceptron neural networks; several combinations were tested between two activation functions of the intermediate layer (hidden layer) and the number of neurons in this layer. The best fitting results in the artificial neural network modeling were found with the sigmoid activation function and three neurons in the hidden layer ([R.sup.2] = 0.990 and RMSE = 2.855 in the training phase). Considering the errors (RMSE, MAE, and MAPE) and the coefficient of determination as criteria for best fit, the artificial neural network models better estimated the leaf area in the training and validation phases. Therefore, the artificial neural network methodology can be used as alternative for estimating leaf area of coffee plants. Keywords: statistical models, artificial intelligence, backpropagation, leaf length and width Dados de area foliar de plantas de cafe sao importantes para estudos e analises de produtividade, fisiologia, adaptacao as condicoes ambientais e manejos culturais. O objetivo deste trabalho foi predizer a area foliar de plantas de cafe por meio de redes neurais artificiais e avaliar a eficiencia dessa metodologia por meio de comparacao com modelos de regressao multipla. Foram avaliados 43 genotipos de reproducao e idade semelhantes e testados 14 tipos de equacoes de regressao multipla a partir de combinacoes de comprimento e largura de folhas O algoritmo backpropagation foi utilizado para desenvolver redes neurais do tipo perceptron multicamadas, e foram testadas possiveis combinacoes entre duas funcoes de ativacao da camada intermediaria e o numero de neuronios na camada intermediaria. Na modelagem de redes neurais artificiais, os melhores resultados de ajuste foram encontrados com a funcao de ativacao sigmoide e tres neuronios na camada oculta ([R.sup.2] = 0,990; RMSE = 2,855 na fase de treinamento). Considerando os erros (RMSE, MAE e MAPE) e coeficientes de determinacao como parametros