Simultaneous estimation of odor classes and concentrations using an electronic nose with function approximation model ensembles

摘要: This paper sets up a practical electronic nose for simultaneously estimating many kinds of odor classes and concentrations. Mathematically, such simultaneous estimation problems can be regarded as multi-input/multi-output (MIMO) function approximation problems. After decomposing an MIMO task into multiple many-to-one tasks, we use model ensembles to implement them one after another. A single may multivariate logarithmic regression, quadratic multilayer perceptron, or support vector machine. An ensemble is made the above four models, represents special kind odor, realizes relationship between sensor array responses represented Naturally, all members in are trained only by samples from odor. The real outputs average predicted concentrations relative standard deviations (R.S.D.s). with smallest R.S.D. finally gives label concentration sample, which looked upon minimum combination rules. results fragrant materials, ethanol, ethyl acetate, caproate, lactate, 21 total, show that proposed strategies effective