Early warning in egg production curves from commercial hens: A SVM approach

Abstract

Early warning in egg production curves from commercial hens: A SVM approach. AbstractArtificial Intelligence allows the improvement of our daily life, for instance, speech and handwritten text recognition, real time translation and weather forecasting are common used applications. In the livestock sector, machine learning algorithms have the potential for early detection and warning of problems, which represents a significant milestone in the poultry industry. Production problems generate economic loss that could be avoided by acting in a timely manner.In the current study, training and testing of support vector machines are addressed, for an early detection of problems in the production curve of commercial eggs, using farm´s egg production data of 478,919 laying hens grouped in 24 flocks.Experiments using support vector machines with a 5 k-fold cross-validation were performed at different previous time intervals, to alert with up to 5 days of forecasting interval, whether a flock will experience a problem in production curve. Performance metrics such as accuracy, specificity, sensitivity, and positive predictive value were evaluated, reaching 0-day values of 0.9874, 0.9876, 0.9783 and 0.6518 respectively on unseen data (test-set).The optimal forecasting interval was from zero to three days, performance metrics decreases as the forecasting interval is increased. It should be emphasized that this technique was able to issue an alert a day in advance, achieving an accuracy of 0.9854, a specificity of 0.9865, a sensitivity of 0.9333 and a positive predictive value of 0.6135. This novel application embedded in a computer system of poultry management is able to provide significant improvements in early detection and warning of problems related to the production curve.

Automated early detection of drops in commercial egg production using neural networks

“…The resulting PPV was 0.8125 and 0.7564 for the 0-d and 1-d prediction intervals, respectively, which showed improvements of more than 23% compared to the model published by Ramírez-Morales et al (2016).…”

“…The window size and the feature selection threshold, which are far from ideal, display a lower performance, which may be related to insufficient information contained in the selected features or excessive and noisy information for the model. In the model developed by Ramírez-Morales et al (2016), multiples of 7 for the window size values were required because of the method used by the expert to calculate these features. In the model proposed here, the information contained in the selected relevant features allows for better modelling of the problem without the restriction mentioned before.…”

“…For the proposed model, the chosen features correspond to the original raw data, which confer less complexity compared to the required features by Ramírez-Morales et al (2016). Variables, such as the age of the birds, cumulative mortality, environmental temperature and humidity, were not used as inputs to the model because of their lack of importance according to the automatic FS technique.…”

“…Xiao et al (2011) performed a statistical analysis of the production curve to identify possible problems. More recently, Ramírez-Morales et al (2016) have shown the suitability of support vector machines with a sliding window for the development of early warning models.…”

Aprendizaje de máquina y aprendizaje profundo en biotecnología: aplicaciones, impactos y desafíos

“…Imágenes de las plantas obtenidas específicamente para este estudio SVM (Morales, Cebrián, Fernandez-Blanco, & Sierra, 2016) Este estudio consiguió desarrollar un predictor para la detección temprana de problemas en la curva de producción de las gallinas con de la 98 % de precisión.…”