期刊
CHEMOMETRICS AND INTELLIGENT LABORATORY SYSTEMS
卷 237, 期 -, 页码 -出版社
ELSEVIER
DOI: 10.1016/j.chemolab.2023.104808
关键词
Sensor; Machine learning; Chemical property; Feature; Classification performance
Chemical recognition using machine learning based on gas sensor signals can be inaccurate when relying solely on signal data. To address this, we propose a novel framework that incorporates predicted chemical properties of analytes to improve classification performance. Experiments with gas sensor array datasets showed improved gas species classification by combining raw features with predicted chemical property features. Our framework bridges the gap between gas sensor signals and target analytes, enhancing classification beyond models trained only on sensor response data.
Chemical recognition using machine learning based on detection by gas sensors relies on the accuracy and sensitivity of the sensors at capturing the key features of target classes. In some cases, however, the electronic signal transduced from the detection of analytes does not completely represent the key attributes, resulting in inaccurate classification results when trained from signal data alone. To overcome this shortcoming, we propose a novel chemistry-informed machine learning framework composed of two modules. From available sensor response data, Module 1 identifies and predicts the chemical properties of the analytes that give rise to the sensitivity and selectivity of the sensors, and Module 2 performs final classifications using the dataset concat-enating predicted chemical properties and raw sensor responses. To evaluate the performance and generaliz-ability of our methodology, we conducted experiments with three gas sensor array datasets for gas detection. In all the cases, the performance of gas species classification was improved when the raw features were combined with the predicted chemical property features. The main contribution of our framework is that it bridges the gap between the gas sensor signals and the target analytes, thereby improving classification performance beyond that of models trained exclusively on sensor response data.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据