A bio-inspired olfactory model using hierarchical temporal memory

Abstract

Artificial olfactory information recognition is a challenging task for signal processing and machine learning. In this paper we explore a new olfactory information recognition method to process the signals from a chemical sensor array of electronic nose, which based on a technology modeling the human brain known as hierarchical temporal memory (HTM). HTM is a broad paradigm for pattern recognition and forward prediction that exploits the hierarchy in time and space existing in the physical world during both learning and inference. In this paper we focus on HTM's capabilities for pattern recognition. We applied HTM to an electronic nose system to discriminate four typical volatile organic compounds (VOCs). Feature vectors were input into the HTM learning algorithm, which were extracted from the sensor array response data, to investigate its generalization capability for olfactory information process. In comparison with the traditional pattern recognition method support vector machine, our experimental results show that the HTM algorithm has a good performance in classification of these VOCs. Its robust generalization capability is suitable for electronic nose applications to process the time series signals from sensor arrays.