A hybrid classifier combination for home automation using EEG signals
- PDF / 1,019,505 Bytes
- 13 Pages / 595.276 x 790.866 pts Page_size
- 52 Downloads / 238 Views
(0123456789().,-volV) (0123456789().,-volV)
S.I. : APPLYING ARTIFICIAL INTELLIGENCE TO THE INTERNET OF THINGS
A hybrid classifier combination for home automation using EEG signals Partha Pratim Roy1
•
Pradeep Kumar1 • Victor Chang2
Received: 13 November 2019 / Accepted: 19 February 2020 Springer-Verlag London Ltd., part of Springer Nature 2020
Abstract Over the years, the usage of artificial intelligence (AI) algorithms is increased to develop various smart applications using Internet-of-Things. Home automation is a fast emerging area that involves monitoring and controlling of household appliances for user comfort and efficient management. Using mental commands to control different electrical appliances and objects in house is a very interesting application. Brain–Computer Interface is used to relay the information from the subject’s brain to an Electronic device, and such devices can be used for this purpose. The information from the subject’s brain is collected in form of Electroencephalogram (EEG) signals. In this paper, we analyze the use of EEG signals for applications related to home automation. We present a hybrid model which makes use of Long Short-Term Memory which is considered as a robust temporal classification model in AI and classical Random Forest Classifier for EEG classification. We also discuss how our proposed hybrid model overcomes the limitation presented by the individual models. To arrive at the best model, we have analyzed various parameters such as sampling rate and combination of different brain rhythms which we finally use in our hybrid model. Based on experiments conducted on a custom-built dataset, we also discuss the spatial significance of different electrodes of the EEG device and get insight in signals generated from different areas of the brain. Keywords EEG IoT LSTMs Hybrid machine learning
1 Introduction The increasing popularity of artificial intelligence (AI) algorithms leads to the development of various smart applications using Internet-of-Things (IoT) devices. These IoT devices are capable of transferring information over the internet by integrating data from various devices to provide different solutions. Home automation is remote monitoring and controlling of electrical appliances. Currently, remote monitoring is done manually using remote
& Partha Pratim Roy [email protected] Pradeep Kumar [email protected] Victor Chang [email protected] 1
Department of Computer Science and Engineering, IIT Roorkee, Roorkee 247667, India
2
School of Computing, Engineering and Digital Technologies, Teesside University, Middlesbrough, UK
controller, voice commands, gestures, etc. What if we could do remote controlling simply through mental commands, i.e., merely through our thoughts? This will help people do mundane tasks easily in their hectic life. In this paper, we present a prototype to test the feasibility of such a model. This prototype has only one appliance, bulb, which we are trying to switch on/off by mental commands. The rise of modest off-the-shelf
Data Loading...
