On the Use of the SVM Approach in Analyzing an Electronic Nose Manlio Gaudioso gaudioso@deis.unical.it. Walaa Khalaf 1 walaa@deis.unical.it. Calogero Pace cpace@unical.it. Dipartimento di Elettronica Informatica e Sistemistica, Universit` a della Calabria, 87036 Rende (CS), Italia Abstract We present an Electronic Nose (ENose) which is aimed both at identifying the type of gas and at estimating its con- centration. Our system contains 8 sensors, 5 of them being gas sensors (of the class TGS from FIGARO USA, INC., whose sensing element is a tin dioxide (SnO 2 ) semiconduc- tor), the remaining being a temperature sensor (LM35 from National Semiconductor Corporation), a humidity sensor (HIH–3610 from Honeywell), and a pressure sensor (XFAM from Fujikura Ltd.). Our integrated hardware–software system uses some ma- chine learning principles and least square regression prin- ciple to identify at first a new gas sample, and then to esti- mate its concentration, respectively. In particular we adopt a training model using the Support Vector Machine (SVM) approach to teach the system how discriminate among dif- ferent gases, then we apply another training model using the least square regression, for each type of gas, to predict its concentration. 1. Introduction The paper deals with the problems of gases detection and recognition as well as with the estimation of their concen- trations. In fact, detection and recognition can be seen as a two–class and a multi–class classification problem, respec- tively. The detection of volatile organic compounds (VOCs) has become a serious task in many fields, because the fast evaporation rate and toxic nature of VOCs could be danger- ous at high concentration levels in air and working ambients for the health of human beings. In fact, the VOCs are also considered as the main reason for allergic pathologies, skin and lung diseases. In this paper, to identify the type of gas we use the sup- 1 Corresponding author. port vector machine (SVM) approach which was introduced by Vapnik [11] as a classification tool. The SVM method strongly relies on statistical learning theory. Classification is based on the idea of finding the best separating hyper- plane (in terms of classification error and separation mar- gin) of two point–sets in the sample space (which in our case is the Euclidean eight–dimension vector space). Our classification approach includes the possibility of adopting Kernel transformations within the SVM context. We adopt a multi–sensor scheme and useful information is gathered by combining the outputs of the different sensors. The use of just one sensor does not allow in general to iden- tify the gas. In fact the same sensor output may correspond to different concentrations of many different gases. On the other hand by combining the information coming from sev- eral sensors of diverse types we identify the gas and esti- mate its concentration. In this paper we present the descrip- tion of the system as shown in Figure 1 producing the details of its construction, a brief theoretical overview of the math- ematical models for classification and regression, and the results of our experiments on four different types of gases (Methanol, Ethanol, Acetone, and Benzene). The results have been particularly encouraging both in terms of clas- sification errors and concentration prediction. In particular we present in section 2 the scheme of the electronic nose (ENose), in section 3 we give a brief overview of the SVM approach. Section 4 is devoted to the description of the ex- periments and results. Finally the conclusions are drawn in section 5. 2. Electronic nose An electronic nose combines an array of gas sensors, whose response constitutes an odor pattern [9]. A single sensor in the array should not be highly specific in its re- sponse but should respond to a broad range of compounds, so that different patterns are expected to be related to dif- Seventh International Conference on Hybrid Intelligent Systems 0-7695-2946-1/07 $25.00 © 2007 IEEE DOI 10.1109/HIS.2007.16 42