Modeling the effect of warehouse hygiene on the population of date-stored pests using an artificial neural network

This study investigated the effects of common date-storing indicators on the frequency of store pests in six date-producing provinces of Iran in 2022. The research employs an artificial neural network model with three layers: input, hidden, and output. Each layer contains a group of nerve cells which are generally related to all the neurons of other layers, including compliance and non-compliance with the storage index and the unobservable factor resulting from factor analysis. The results showed that the indicators of product layout, including pallets, storage containers, cultivars separation, spacing, and shelving, respectively, have the greatest effect on the severity of O. surinaemensis, E. elutella, E. elutella, E. elutella, and E. calidella. Building indicators, including flooring, internal walls, sloping passage at the entrance, proper access to internal spaces, and the presence of a dock have the greatest effect on the contamination of D. melanogaster, D. melanogaster, E. kuhniella, O. surinaemensis, and E. kuhniella. The indicators of setting the environmental conditions of the warehouse, including ventilation systems, temperature, humidity, environmental hygiene, and warehouse floor, respectively, had the greatest effect on the contamination of E. kuhniella, D. melanogaster, D. melanogaster, E. kuhniella, and D. melanogaster. The neural network model included 28 synapses between different layers. The lighting system, proper access, spacing, and flooring are effectively predict warehouse pests’ population density by 100%, 94.9%, 77.1%, and 60.2%, respectively. These factors are crucial in pest infestations. The algorithm obtained from the neural network determined the importance of complying with storage standards in managing date pests.