Prevention of Cooktop Ignition Using Detection and Multi-Step Machine Learning Algorithms
Wai Cheong Tam, Eugene Yujun Fu, Amy E. Mensch, Anthony P. Hamins, Christina Yu, Grace Ngai, Hong va Leong
This paper presents a study to examine the potential use of machine learning models to build a real-time detection algorithm for prevention of unattended cooking fires. 16 sets of time- dependent sensor signals were obtained from 60 normal/ignition cooking experiments. A total of 200,000 data instances are documented and made available in the public domain. The raw data are preprocessed. Hand-crafted features for time series data focusing on real-time detection application are provided. Utilizing the leave-one-out cross validation method, three machine learning models are built and tested. Parametric studies are carried out to understand the diversity, volume, and tendency of the data. Given the current dataset, the detection algorithm based on Support Vector Machine provides the most reliable prediction (with an overall accuracy of 96.9 %) on hazardous conditions due to unattended cooking. Additional analysis demonstrates that using a multi-step approach can further improve the overall prediction accuracy. This work will contribute to the development of an accurate detection algorithm which can provide reliable feedback to intercept ignition of unattended cooking and help reduce fire losses from cooking fires.
April 27-May 1, 2020
International Association of Fire Safety Science 2020