Chaos based neural network optimization for concentration estimation of indoor air contaminants by an electronic nose

Reservoir computing compensates slow response of chemosensor arrays exposed to fast varying gas concentrations in continuous monitoring

“…However, the capacity of memory in tappeddelayed networks is limited by the duration of the defined delay. Simple linear models only consider the value of the sensor after it reached the steady-state [43,44,45,46]. Such a feature extraction procedure assumes that the gas composition remains constant until the sensor reaches the steady-state and discards any information contained in the sensor's dynamics.…”

A novel sensor feature extraction based on kernel entropy component analysis for discrimination of indoor air contaminants

“…In previous work [24][25][26][27], the E-nose system has been presented. The E-nose consists of pattern recognition and metal oxide semiconductor gas sensors array, in which five different types of gas sensors including four TGS sensors (i.e.…”

A rapid discreteness correction scheme for reproducibility enhancement among a batch of MOS gas sensors

“…Metal oxide semiconductor (MOS) gas sensors have been widely reported in E-nose for detection of chemicals [3][4][5][6][7][8][9][10][11][12]. MOS sensors have also been used in odor-discrimination system for fruit detection [13].…”