Figure 10
Two simplified neural network topology diagrams illustrate models for predicting user satisfaction in buildings.The models are titled “Apartments - N N Model Topology” and “Offices - N N Model Topology.” Both share the same architecture: an “Input Layer” (yellow) with seven nodes representing user perception and satisfaction features: “Thermal Sensation,” “Thermal Satisfaction,” “Sound Perception,” “Sound Satisfaction,” “Daylight Perception,” “Daylight Satisfaction,” and “Air Quality Perception.” Each model includes two “Hidden Layers” (light blue): the first with 100 nodes and the second with 60 nodes. The structure between the two hidden layers is not explicitly labeled. An “Output Layer” (orange) contains three nodes representing the predicted satisfaction levels: “Satisfied,” “Neutral,” and “Dissatisfied.” All nodes in each layer are fully interconnected.

NN simplified topologies diagrams for apartment NN model (top) and office NN model (bottom). Source: Authors' own work

or Create an Account

Close subscription notice
Close access options