| Title |
A Study on Prediction Method for Preferred Dimming Level of Lighting User Using Usage Environment Data |
| Authors |
Jai-Won Chung ; Sangsu Park ; Seok-Oh Bang ; Seong-Sik Yoo ; Hyun-Seok Choi ; Jin-Sung Rho |
| DOI |
http://doi.org/10.5207/JIEIE.2024.38.1.001 |
| Keywords |
Classification model; Dimming level; LED lighting; Random forest; Usage environment data |
| Abstract |
Few studies have looked at the prediction method for preferred dimming level of lighting user using usage environment data(usage time, weather, etc.) and machine learning methods. In this paper, we propose the random forest model for prediction of preferred dimming level of lighting user using usage environment data. Temperature, relative humidity, precipitation, wind speed, hour, AM/PM status, day of the week, weekday/ weekend status, and dimming setting were measured at one hour interval through the experiment, and the relation of each variable were investigated through the PhiK correlation coefficient. About 70% of the total data (n = 171) was used as the training set data (n = 119) for construction of the random forest model. The model using the 3-fold cross validation method were constructed using the training set data. About 30% of the total data was used as the test set data (n = 52) for evaluation of the model. The accuracy of random forest model using two values(hour, temperature) were 78.9%. |