الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 84 |  |  |
| | | from | 84 | | |
Abstract
Food quality analysis and measurement is considered a difficult task. NIR spectroscopy is frequently used to examine the chemical and physical characteristics of the material in many different domains. Our goal is to offer an easy method for evaluating the quality of food and for measuring the concentration of a particular constitute of a specific food item. We used spectrometers built with MEMS to get our data. We introduce an IoT application that focuses on food quality in this study. We employ the spectrometer as a general-purpose sensor to produce a Near-infrared (NIR) spectra that can be employed to collect NIR spectra, and then we use these spectra to create regression models that target milk quality. For several milk constituents, including fat, protein, lactose, and solids-non-fat, we solve a regression problem. We use a number of comparison regression model combinations, starting from simple linear machine learning models to sophisticated Deep Learning methods. It is shown that the Feed Forward Neural Networks gives best performance for the four milk constitutes we are studying.