An Online Data Validation Algorithm for Electronic Nose
An electronic nose (e-nose) is a device that analyzes the chemical components of an odour. The e-nose consists of an array of gas sensors for chemical detection, and a mechanism for pattern recognition to return the odour concentration. Odour concentration defines the identifiability and perceivability of an odour. Given that accurate prediction of odour concentration requires valid measurements, automatic assessment of sampled measurements is of prime importance. The impairment of the e-nose, and environmental factors (including wind, humidity, temperature, etc.) may introduce significant amount of noise. Inevitably, the pattern recognition results are affected. We propose an online algorithm to evaluate the validity of sensor measurements during the sampling before using the data for pattern recognition phase. The proposed algorithm is computationally efficient and straightforward to implement.
KeywordsArtificial olfaction Computational complexity Electronic nose Gas sensor Outlier detection Robust covariance estimation
The project was funded by the natural sciences and engineering research council of Canada (NSERC) through the industrial partnership Engage program. Vahid Partovi Nia is partially supported by the Canada excellence research chair in data science for real-time decision making.
- 3.Bermak, A., Belhouari, S.B., Shi, M., Martinez, D.: Pattern recognition techniques for odor discrimination in gas sensor array. Encycl. Sens. X, 1–17 (2006)Google Scholar
- 5.Carlo, S.D., Falasconi, M.: Drift correction methods for gas chemical sensors in artificial olfaction systems: techniques and challenges. Adv. Chem. Sens. 14, 305–326 (2012)Google Scholar
- 9.Donoho, D.L.: Breakdown properties of multivariate location estimators. Ph.D. Qualifying Paper Harvard University (1982)Google Scholar
- 12.Gardner, J., Bartlett, P.: A brief history of electronic noses. Sens. Actuators B: Chem. 18, 211–220 (1994)Google Scholar
- 16.Holmes, M.P., Gray, A.G., Isbell, C.L.: Fast SVD for large-scale matrices. In: Workshop on Efficient Machine Learning at NIPS, vol. 58 (2007)Google Scholar
- 25.McGinley, P.C., Inc, S.: Standardized odor measurement practices for air quality testing. In: Air and Waste Management Association Symposium on Air Quality Measurement Methods and Technology, San Francisco, CA (2002)Google Scholar
- 26.Mirshahi, M., Partovi Nia, V., Adjengue, L.: Statistical measurement validation with application to electronic nose technology. In: Proceedings of the 5th International Conference on Pattern Recognition Applications and Methods, pp. 407–414 (2016)Google Scholar
- 32.Stahel, W.A.: Robust estimation: infinitesimal optimality and covariance matrix estimators. Ph.D. thesis, ETH, Zurich (1981)Google Scholar