Abstract
This paper presents a stacked sparse auto-encoder (SSAE) based deep learning method for an electronic nose (e-nose) system to classify different brands of Chinese liquors. It is well known that preprocessing; feature extraction (generation and reduction) are necessary steps in traditional data-processing methods for e-noses. However, these steps are complicated and empirical because there is no uniform rule for choosing appropriate methods from many different options. The main advantage of SSAE is that it can automatically learn features from the original sensor data without the steps of preprocessing and feature extraction; which can greatly simplify data processing procedures for e-noses. To identify different brands of Chinese liquors; an SSAE based multi-layer back propagation neural network (BPNN) is constructed. Seven kinds of strong-flavor Chinese liquors were selected for a self-designed e-nose to test the performance of the proposed method. Experimental results show that the proposed method outperforms the traditional methods.
Highlights
In recent years, with the improvement of people’s living standard, people pay more and more attention to food safety
We present a stacked sparse auto-encoder (SSAE) [19] based deep learning (DL) approach for Chinese liquors classification
Considering that the samples number is small, we adopted the way of cross-validation to evaluate the performance of the proposed SSAE-back propagation neural network (BPNN) and the traditional methods, where the cross-validation could Considering eliminatethat the the over-fitting
Summary
With the improvement of people’s living standard, people pay more and more attention to food safety. Jia et al [6] proposed a new hybrid algorithm for Chinese liquors classification, in which man-made features were reduced using a combined KECA-LDA technique and the extreme learning machine (ELM) was applied as a classifier. Before they found this optimal KECA-LDA-ELM combination algorithm, they have tried some other combination algorithms, such as KECA-BPNN, KECA-ELM, and KECA-LDA-BPNN. The remainder of this paper is organized as follows: Section 2 presents the method description, including spare auto-encoder, SSAE based BPNN method and traditional methods applied to the data processing of the e-nose we designed.
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call