Data Analysis for Chemical Sensor Arrays

Abstract

Arrays were introduced in the mid-eighties as a method to counteract the cross-selectivity of gas sensors. Their use has since become a common practice in sensor applications. [1]. The great advantage of this technique is that once arrays are matched with proper multivariate data analysis, the use of non-selective sensors for practical applications becomes possible. Again in the eighties, Persaud and Dodds argued that such arrays has a very close connection with mammalian olfaction systems. This conjecture opened the way to the advent of electronic noses [2], a popular name for chemical sensor arrays used for qualitative analysis of complex samples.It is worth remarking that a gas sensor array is a mere mathematical construction where the sensor outputs are arranged as components of a vector. Arrays can also be utilized to investigate the properties of chemical sensors, or even better, the peculiar behaviour of a sensor as a component of an array. In this chapter, the more common sensor array methodologies are critically reviewed, including the most general steps of a multivariate data analysis. The application of such methods to the study of sensor properties is also illustrated through a practical example.

Methods for Early Detection of Microbiological Infestation of Buildings Based on Gas Sensor Technologies

“…Therefore, for chemical sensors, feature extraction evaluates certain parameters from a signal stream, which represent the information pertaining to the objective of classification. It is of key significance, since it governs the sensor output used in estimating the measured quantities [153].…”

“…The measurement data from sensor arrays are multidimensional; therefore, it is difficult to derive important information on the investigated gas sample without the application of appropriate analysis methods. In a dataset pertaining to the number of measurements, the explanatory techniques, employed following the feature extraction, as well as preprocessing and normalization, are used to investigate the characteristics of data and discover its internal properties [153]. The exploratory analysis of data indicates the adequateness of the sensor array for a given task; then, supervised classification can be carried out to build a model for predicting the class membership of the investigated samples.…”

Principal Component Analysis – A Realization of Classification Success in Multi Sensor Data Fusion

“…The ultimate task of these sensors is to collect the digital fingerprint or signals that would be further interpreted using multivariate statistical tools before the objective of the fusion approach is attained. One of the most popular exploratory data analyses in chemical sensors is PCA (Di Natale et al, 2006). PCA is a procedure that permits to extract useful information from the data, to explore the data structure, the relationship between the objects and features, and the global correlation of the features.…”

Statistical Analysis of Chemical Sensor Data

“…Similar to other high throughput biological experiments, several initial steps are often necessary before analyzing the data for a scientific question. Natale et al (2006) discuss several preprocessing steps including feature extraction, zero-centered scaling, autoscaling, and normalization. Jurs et al (2000) outline these techniques for chemical sensor arrays, and further include discussion on background or baseline subtraction, and linearization.…”

“…Linearization can also be performed by considering the engineering-derived equations that drive the signal (see, e.g., Robins et al (2005)). Some chemical sensors are ruled by a power law relationship between sensor signal and analyte concentration; this is often the case, for example, with metal-oxide semiconductor gas sensors (Natale et al;. Using least squares approaches, it is possible to estimate the parameters in a power law relationship.…”