Using an artificial neural network to predict the probability of oviposition events of precision-fed broiler breeder hens

You, J., E. Lou, M. Afrouziyeh, N. M. Zukiwsky, and M. J. Zuidhof. 2021. Using an artificial neural network to predict the probability of oviposition events of precision-fed broiler breeder hens. Poultry Science 100:101187. doi https://doi.org/10.1016/j.psj.2021.101187

PIP Author(s)

Categories

feed restriction PIP Newsletter maintenance Small Flock Poultry Course Uncategorized feather gut microbiota News broiler Precision Feeding broiler breeder Research Showcase 2021 Gompertz model flock talk water purification Keratin biopolymers egg yolk antioxidants faba bean bone carcass Research Showcase 2022 Event Proceedings Spent hen Research Showcase 2023 egg white peer-reviewed Prairie Poultry Meeting 2020 Applied Poultry Science Webinar 2020 laying hen Collagen growth model chicken Presentation phytase egg growth performance Publication People of Poultry energy partitioning model Pathogens & Protection Innovation Showcase sexual maturity Inflammation ACP webinars

About

Publication Metrics

Identifying daily oviposition events for individual broiler breeders is important for bird management. Identifying non-laying birds in a flock that might be caused by improper nutrition or diseases can guide diet changes or disease treatments for these individuals. The day and time of oviposition events of individual birds in a free-run flock can be hard to predict. Based on a precision feeding (PF) system that can record the feeding activity of individual birds, a recent study reported a machine learning model to predict daily egg-laying events of broiler breeders. The objective of the current study was to improve the previous approach in 2 aspects: 1) To apply the model on the current day; 2) To output more informative results. To accomplish this, an artificial neural network (ANN) model was built to predict the probability of oviposition events occurring.

Approach

In this study, data were obtained from a flock of broiler breeders (n = 95) raised in 2 environmentally controlled chambers. Each chamber was equipped with 2 PF stations. After photo-stimulation at 22 wk of age, the egg production of individual hens was recorded on a daily basis. If a hen laid an egg in 1 day, it was considered as an egg-laying event; if a hen did not lay an egg in 1 day, it was considered as a no-egg-laying event. A traditional trap nest box with 8 nesting sites was placed in each pen, and it was checked every hour from 07:30 to 17:30 every day. During the study, 706 egg-laying events occurred in the RFID nest box, while the remaining egg-laying events occurred in the trap nest box. The total number of no-egg-laying events was 3,559. Python 3.7.0 was used to facilitate data preprocessing, feature engineering, and model construction. The ANN model was evaluated by the receiver operating characteristic (ROC) curve and the area under the ROC curve (AUC).

Analysis of Results

AUC was 0.9409, which meant a 94.09% chance to correctly distinguish an egg-laying event from a no-egg-laying event. The ANN model could predict oviposition events on the current day, and the output was a probability that could be informative to indicate the likelihood of an oviposition event for an individual breeder. In situations where total egg production was known for a group, the ANN model could predict the probability of daily oviposition events occurring of all individual birds and then rank them to choose those most likely to have laid an egg.

Application

The AUC value of the ANN model was 0.9409, indicating the ANN model had an outstanding classification performance. A possible application scenario of using the ANN model was to identify the breeders that have laid an egg in the pen. The ANN model could be used to predict oviposition events before the end of each day, and the prediction outputs were informative probabilities that indicated the likelihood of oviposition by individual hens within each day. In situations where the total egg production for a flock of breeders in one day was known, the probability of daily oviposition events occurring of all individual birds could be predicted and then ranked to choose those most likely to have laid an egg.

Abstract

Identifying daily oviposition events for individual broiler breeders is important for bird management. Identifying non-laying birds in a flock that might be caused by improper nutrition or diseases can guide diet changes or disease treatments for these individuals. Oviposition depends on follicle maturation and egg formation, and follicle maturation can be variable. As such, the day and time of oviposition events of individual birds in a free-run flock can be hard to predict. Based on a precision feeding (PF) system that can record the feeding activity of individual birds, a recent study reported a machine learning model to predict daily egg-laying events of broiler breeders. However, there were 2 limitations in that study: 1) It could only be used to identify daily egg-laying events on a subsequent day; 2) The prediction outputs that were binary labels were unable to indicate more details among the outputs with the same label. To improve the previous approach, the current study aimed to predict and output the probability of daily oviposition events occurring using a specific time point in 1 day. In the current study, 706 egg-laying events recorded by nest boxes with radio frequency identification of hens and 706 randomly selected no-egg-laying events were used. The anchor point was newly defined in the current study as a specific time point in 1 day, and 26 features around the anchor point were created for all events (706 egg-laying events and 706 no-egg-laying events). A feed-forward artificial neural network (ANN) model was built for prediction. The area under the receiver operating characteristic (ROC) curve was 0.9409, indicating that the model had an outstanding classification performance. The ANN model could predict oviposition events on the current day, and the output was a probability that could be informative to indicate the likelihood of an oviposition event for an individual breeder. In situations where total egg production was known for a group, the ANN model could predict the probability of daily oviposition events occurring of all individual birds and then rank them to choose those most likely to have laid an egg.