Outdoor clothing choice for different populations in cold regions: A clothing choice prediction model based on machine learning

The choice of clothing is one of the main behavioral ways for people to adjust their thermal comfort outdoors. However, the choice of clothing is influenced by a combination of factors. The researchers need to use a method that deals with multivariable nonlinear problems to predict clothing insulation for outdoor people. Therefore, this study proposes an outdoor clothing prediction model based on the gray wolf optimization algorithm backpropagation neural network.This study starts with conditional assumptions. The clothing choice of outdoor residents is influenced by various climate parameters and gender. In a field experiment, outdoor environmental parameters were measured in real time, and respondents were surveyed using subjective questionnaires. The experimental data in the field and validation tests of the model confirmed this hypothesis. A total of 2883 valid samples were collected in this study, and an outdoor clothing prediction model that considered multiple nonlinear comprehensive effects was established. The results show that the accuracy of the proposed model is considerably higher than that of the traditional linear model. For the four sample groups, the root mean square prediction errors of the model were 9.57%, 2.96%, 4.87%, and 6.96%. The machine learning based outdoor clothing prediction model established in this paper not only provides a theoretical basis for outdoor thermal comfort evaluation but also provides a reference for research in other areas.