Using machine learning technique for disease outbreak prediction in rainbow trout (Oncorhynchus mykiss) farms

YILMAZ M., Çakir M., ORAL O., Oral M. A., Arslan T.

Aquaculture Research, cilt.53, sa.18, ss.6721-6732, 2022 (SCI-Expanded) identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 53 Sayı: 18
  • Basım Tarihi: 2022
  • Doi Numarası: 10.1111/are.16140
  • Dergi Adı: Aquaculture Research
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Academic Search Premier, Aquatic Science & Fisheries Abstracts (ASFA), Artic & Antarctic Regions, BIOSIS, CAB Abstracts, Environment Index, Food Science & Technology Abstracts, Geobase, INSPEC, Pollution Abstracts, Veterinary Science Database
  • Sayfa Sayıları: ss.6721-6732
  • Anahtar Kelimeler: aquaculture, disease outbreak prediction, machine learning, rainbow trout, sustainability
  • Akdeniz Üniversitesi Adresli: Evet

Özet

© 2022 John Wiley & Sons Ltd.Water quality parameters such as temperature, dissolved oxygen, pH and total dissolved solids are important environmental factors affecting fish welfare. The deterioration of these parameters beyond the tolerance limits causes environmental stress and suppression of the immune system. Moreover, it allows opportunistic pathogens that are always present in the environment to infect immune-suppressed fish and cause serious disease outbreaks. In this study, water quality parameters and pathogenic bacteria profiles were monitored for 1 year in rainbow trout farms operating in the same river basin. Then, a data set was created considering the pathogenic bacteria in the diseased fish and the water quality parameters in the farm environment. Each of the water quality parameters in the data set was first used as an attribute and their order of importance in terms of disease outbreak was determined. Then, using multinomial logistic regression (MLR) analysis, which is one of the machine learning (ML) techniques, the possibility of water quality parameters revealing a disease outbreak was evaluated. Furthermore, very effective models that can be used to predict the probability of disease occurrence in trout farms with an accuracy of 95.65% have been created.