Machine Learning and Interior Design Generation Study on Residential Preferences of Youth Groups
Year: 2026
Type: conference-paper
Abstract: Aiming at the problems of poor flexibility of household type, low efficiency of youth apartment design cycle, irrational household type design, poor tenant living experience and deviation of youth group's living preferences, etc., we construct a triangular cross-modeling research framework based on case study, machine learning method, and user preference research, and based on the analysis of user preference, we train the dataset through machine learning method and apply it to the design practice of space generation. space generation design practice, aiming to reduce designers' repetitive work and improve the efficiency of design work output, and optimize the workflow of youth apartment design for different people's needs. This study draws the following two conclusions: 1. The data model based on the selection of population living preferences provides a more effective learning dataset for machine learning and the direction of interior generative design; 2. After model training of the dataset, the constraints of the machine learning algorithms are utilized to generate a more controllable design scheme, and the work method of grounded optimization is carried out, so as to expand the boundaries of machine learning in the study of user preferences. This not only improves the adaptability of the interior space design of youth apartments, but also provides a path reference for multidisciplinary crossover and integration research, and lays down the data support for the accurate identification of crowd preference and personalized space design in the future.
Language: English
Subfield: Architecture
Field: Engineering
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W7201857481