Detection of Adulteration in Infant Formula Based on Ensemble Convolutional Neural Network and Near-Infrared Spectroscopy

Abstract

Adulteration in dairy products has received world-wide attention, and at the same time, near infrared (NIR) spectroscopy has proven to be a promising tool for adulteration detection given its advantages of real-time response and non-destructive analysis. Regardless, the accurate and robust NIR model for adulteration detection is hard to achieve in practice. Convolutional neural network (CNN), as a promising deep learning architecture, is difficult to apply to such chemometrics tasks due to the high risk of overfitting, despite the breakthroughs it has made in other fields. In this paper, the ensemble learning method based on CNN estimators was developed to address the overfitting and random initialization problems of CNN and applied to the determination of two infant formula adulterants, namely hydrolyzed leather protein (HLP) and melamine. Moreover, a probabilistic wavelength selection method based on the attention mechanism was proposed for the purpose of finding the best trade-off between the accuracy and the diversity of the sub-models in ensemble learning. The overall results demonstrate that the proposed method yielded superiority regression performance over the comparison methods for both studied data sets, and determination coefficients (R2) of 0.961 and 0.995 were obtained for the HLP and the melamine data sets, respectively.

Prediction of Metabolic Characteristics of Cardiovascular and Cerebrovascular Diseases Based on Convolutional Neural Network

“…Artificial neural network (ANN) is a kind of neural network that simulates the process from activation to signal sending of brain neurons and organizes many neurons according to certain hierarchical structure to form multilayer neural network [13,14] There is a fixed time delay between the input and output of neurons, mainly due to synaptic delay. It can be seen from the above analysis that the MP model, as one of the convolution neural network models, can calculate different types of data.…”

“…Artificial neural network (ANN) is a kind of neural network that simulates the process from activation to signal sending of brain neurons and organizes many neurons according to certain hierarchical structure to form multilayer neural network [ 13 , 14 ]. At first, the neural network in biology was abstracted into a simple linear model based on threshold logic algorithm, called MP (McCulloch-Pitts) neuron model.…”

Hyperspectral detection of walnut protein contents based on improved whale optimized algorithm

“…In recent years, the deep learning, as a state-of-the-art technique, has been extensively applied in the field of hyperspectral image processing, in which the features could be learned automatically according to the targeted tasks [14][15][16][17][18] . And large number of labeled samples is desirable to ensure the stability of the deep learning models.…”

Convolutional Neural Network for Object Detection in Garlic Root Cutting Equipment

“…CNNs use computational models composed of multiple processing layers to learn abstract data representations and are the latest technological advancement in object detection [ 17 ]. An integrated learning approach based on CNN estimators applied to the determination of infant formula adulterants yielded better regression performance [ 18 ]. Regions with Convolution Neural Network features (R-CNNs) were the first two-phase algorithm object detection model [ 19 ].…”