TY - GEN
T1 - Application of a Fuzzy ARTMAP Neural Network for Indoor Air Quality Prediction
AU - De Assis Pedrobon Ferreira, Willian
AU - Grout, Ian
AU - Da Silva, Alexandre Cesar Rodrigues
N1 - Publisher Copyright:
© 2022 IEEE.
PY - 2022
Y1 - 2022
N2 - Indoor air quality monitoring is an important activity to ensure continued health and well-being of citizens living, studying, and working in indoor environments. This practice has been widely developed through the application of low-cost sensors that are able to measure gas concentrations, particulate matter, and other components such as humidity and temperature that affect indoor air quality. Additionally, machine learning algorithms have been applied in the interpretation of sampled environmental data to improve the performance of monitoring systems. This paper proposes the implementation of a fuzzy ARTMAP neural network, which employs the concepts of Adaptive Resonance Theory (ART), to compute the prediction of particulate matter sampled in a domestic bedroom environment. With the application of a specialized online training architecture, the fuzzy ARTMAP network can be a promising alternative to predict particulate matter time series data modeled in sliding windows, obtaining predictions 24-hour ahead with mean absolute error (MAE) ranging here from 0.26 to 7.65.
AB - Indoor air quality monitoring is an important activity to ensure continued health and well-being of citizens living, studying, and working in indoor environments. This practice has been widely developed through the application of low-cost sensors that are able to measure gas concentrations, particulate matter, and other components such as humidity and temperature that affect indoor air quality. Additionally, machine learning algorithms have been applied in the interpretation of sampled environmental data to improve the performance of monitoring systems. This paper proposes the implementation of a fuzzy ARTMAP neural network, which employs the concepts of Adaptive Resonance Theory (ART), to compute the prediction of particulate matter sampled in a domestic bedroom environment. With the application of a specialized online training architecture, the fuzzy ARTMAP network can be a promising alternative to predict particulate matter time series data modeled in sliding windows, obtaining predictions 24-hour ahead with mean absolute error (MAE) ranging here from 0.26 to 7.65.
KW - fuzzy ARTMAP neural network
KW - indoor air quality
KW - online training
KW - particulate matter prediction
UR - http://www.scopus.com/inward/record.url?scp=85128177831&partnerID=8YFLogxK
U2 - 10.1109/iEECON53204.2022.9741563
DO - 10.1109/iEECON53204.2022.9741563
M3 - Conference contribution
AN - SCOPUS:85128177831
T3 - Proceedings of the 2022 International Electrical Engineering Congress, iEECON 2022
BT - Proceedings of the 2022 International Electrical Engineering Congress, iEECON 2022
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2022 International Electrical Engineering Congress, iEECON 2022
Y2 - 9 March 2022 through 11 March 2022
ER -
