Chitin from seafood waste: particle swarm optimization and neural network study for the improved chitinase production

BACKGROUND The current system in the processing of seafood leads to accumulation of many waste products, such as shells, tails, heads and bones. This seafood waste can be exploited for the extraction of chitin, with numerous applications in different fields. Seafood waste treatment can produce some...

Full description

Saved in:
Bibliographic Details
Published in:Journal of chemical technology and biotechnology (1986) 2022-02, Vol.97 (2), p.509-519
Main Authors: Suryawanshi, Nisha, Eswari, J Satya
Format: Article
Language:English
Subjects:
Algorithms
artificial neural network
Artificial neural networks
Bayesian analysis
Bones
Chitin
Chitinase
Fermentation
Multilayers
Neural networks
Optimization
Particle swarm optimization
Polynomials
Regularization
Seafood
seafood waste
Substrates
valuable products
Waste treatment
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:BACKGROUND The current system in the processing of seafood leads to accumulation of many waste products, such as shells, tails, heads and bones. This seafood waste can be exploited for the extraction of chitin, with numerous applications in different fields. Seafood waste treatment can produce some valuable products. For the valorization of chitin, its degradation is an important step that can be achieved using the chitinase enzyme. Interestingly, chitin can also be used as a significant substrate for chitinase production. In this study, chitinase activity was enhanced by optimizing the fermentation medium, and chitin was used as the substrate. The polynomial model obtained by central composite design was employed in a particle swarm optimization algorithm and artificial neural network to optimize the final optimal concentration factors. The optimization results were compared for the better activity of chitinase. From the authors' best knowledge, the optimization of fermentation medium for chitinase production by particle swarm optimization was performed for the first time. RESULTS The highest activity optimized by particle swarm optimization and artificial neural network/ Bayesian regularization algorithm) was 115.8 and 124.78 U L–1, respectively, with the optimized variables. CONCLUSION This study concluded that particle swarm optimization and artificial neural network are the best optimization methods for medium optimization. Among the multilayer feed‐forward algorithms in the artificial neural network, the Bayesian regularization algorithm was useful in optimizing medium components. © 2020 Society of Chemical Industry
ISSN:0268-2575
1097-4660
DOI:10.1002/jctb.6656