Main content area

Using Genetic Algorithm Neural Network on Near Infrared Spectral Data for Ripeness Grading of Oil Palm (Elaeis guineensis Jacq.) Fresh Fruit

Silalahi, Divo Dharma, Reaño, Consorcia E., Lansigan, Felino P., Panopio, Rolando G., Bantayan, Nathaniel C.
Information processing in agriculture 2016 v.3 no.4 pp. 252-261
Elaeis guineensis, algorithms, neural networks, principal component analysis, raw fruit, spectral analysis
Genetic Algorithm Neural Network (GANN) for multi-class was used to predict the ripeness grades of oil palm fresh fruit using Near Infrared (NIR) spectral data. NIR spectral data provide sufficient information about compound structure of samples from the near infrared light that passes through. The variables used in the GANN modeling process were the new variables obtained as a result of dimensional reduction from original NIR spectral data using Principal Component Analysis (PCA). Three statistical measures such as Mean Absolute Error (MAE), Root Mean Squared Error (RMSE) and the percentage (%) of good classification were used to assess adequacy of the GANN model. Based on the results, the GANN model created was precise enough to be used as the model calibration for this multi-class problem.