Neural network modeling and optimization of process parameters for production of chhana cake using genetic algorithm

Mishra H. N., . and Mukhopadhyay S., . and Goswami T. K., . and Majumdar G. C., . Neural network modeling and optimization of process parameters for production of chhana cake using genetic algorithm. pp. 465-475. ISSN 22317546

Chhana cake locally termed chhana podo is a baked traditional dairy product of India. The present study was undertaken for optimization of process parameters pertaining to production of chhana podo. Independent variables namely moisture content of feed-mix: 52.5 - 62.5 (wb) baking temperature: 60 - 180C baking time: 1 - 9 h and height of feed-mix: 1 - 5 cm were selected heuristically and their effect on dependent variables namely hardness whiteness index yellowness index tint of crust and crumb moisture content and expansion ratio of chhana podo were studied. Although quadratic models fitted to responses exhibited relative deviation percent (Rd) ranging from 1.214 to 5.406; lack of fit was significant for all responses except crust yellowness index and crust tint. Neural network modeling was adopted (Rd for training 1.739 Rd for validation 1.845) and relative importance of factors on responses were found. Optimum conditions obtained from genetic algorithm were: moisture content of feed-mix 57.43 (wb) baking temperature 151.4C baking time 4.35 h height of mix 2.9 cm.

Full text available from: Official URL: http://ifrj.upm.edu.my/22%20(02)%202015/(4).pdf

Abstract

Chhana cake locally termed chhana podo is a baked traditional dairy product of India. The present study was undertaken for optimization of process parameters pertaining to production of chhana podo. Independent variables namely moisture content of feed-mix: 52.5 - 62.5 (wb) baking temperature: 60 - 180C baking time: 1 - 9 h and height of feed-mix: 1 - 5 cm were selected heuristically and their effect on dependent variables namely hardness whiteness index yellowness index tint of crust and crumb moisture content and expansion ratio of chhana podo were studied. Although quadratic models fitted to responses exhibited relative deviation percent (Rd) ranging from 1.214 to 5.406; lack of fit was significant for all responses except crust yellowness index and crust tint. Neural network modeling was adopted (Rd for training 1.739 Rd for validation 1.845) and relative importance of factors on responses were found. Optimum conditions obtained from genetic algorithm were: moisture content of feed-mix 57.43 (wb) baking temperature 151.4C baking time 4.35 h height of mix 2.9 cm.

[error in script]

Item Type:	Article
AGROVOC Term:	Bakery products
AGROVOC Term:	Neural networks
AGROVOC Term:	Optimization methods
AGROVOC Term:	Dairy products
AGROVOC Term:	Quadratic programming
AGROVOC Term:	Cheese
Depositing User:	Ms. Suzila Mohamad Kasim
Last Modified:	24 Apr 2025 06:27
URI:	http://webagris.upm.edu.my/id/eprint/22036

Actions (login required)

View Item