Scholarlink

<p>Advancement of machine learning models hold immense potential in predicting the quality parameters of new food products. Current research study focuses to develop machine learning model capable of accurately predicting the quality parameters of millet flour-infused noodles. Existing research data was used for machine learning model evaluation. Quantity of wheat flour, finger millet flour, pearl millet flour in (g/100&#xa0;g) and noodles quality parameters viz<i>.</i> cooking weight (g/10&#xa0;g), cooking loss (%), firmness and color were identified as crucial input parameters for training machine learning algorithm. The research data were fit into various machine learning algorithms and one with least Mean Square Error (MSE) was selected to predict the quality parameters of noodles prepared across a broad spectrum of millet flour composition. To validate the accuracy of predictive performance of machine learning model, noodles were prepared through real time experiments using the same spectrum of millet flour compositions and quality parameters were measured. Experimental values were on par with the predictive values of all qualities of cooked noodles. The predictive cooking weight of millet flour infused noodles was ranging from 21.5 to 30.0 (g/10&#xa0;g) while the experimental value was ranging from 21.92 to 30.05 (g/10&#xa0;g).</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Machine learning based framework for quality prediction in millet flour-infused noodles

  • 期刊论文

  • 发表日期：

    2025年8月1日

查看全文
  • Vanila Sildas,
  • Manimehalai N.,
  • Rathinam Ananthanarayanan,
  • Rathnakumar K.

摘要

Advancement of machine learning models hold immense potential in predicting the quality parameters of new food products. Current research study focuses to develop machine learning model capable of accurately predicting the quality parameters of millet flour-infused noodles. Existing research data was used for machine learning model evaluation. Quantity of wheat flour, finger millet flour, pearl millet flour in (g/100 g) and noodles quality parameters viz. cooking weight (g/10 g), cooking loss (%), firmness and color were identified as crucial input parameters for training machine learning algorithm. The research data were fit into various machine learning algorithms and one with least Mean Square Error (MSE) was selected to predict the quality parameters of noodles prepared across a broad spectrum of millet flour composition. To validate the accuracy of predictive performance of machine learning model, noodles were prepared through real time experiments using the same spectrum of millet flour compositions and quality parameters were measured. Experimental values were on par with the predictive values of all qualities of cooked noodles. The predictive cooking weight of millet flour infused noodles was ranging from 21.5 to 30.0 (g/10 g) while the experimental value was ranging from 21.92 to 30.05 (g/10 g).