AUTOMATED EVALUATION OF ACOUSTIC QUALITY OF PREMISES USING DEEP LEARNING

This research is dedicated to the development and evaluation of a deep learning model for automated evaluation of acoustic quality of premises. The model was trained on a large dataset, which included various acoustic parameters, and showed high accuracy, sensitivity, specificity, and F1-score. However, some areas were identified where the model could be improved, particularly in assessing the acoustic quality of premises with a high level of noise. Despite these challenges, the research results confirm the potential of using deep learning in the field of acoustic design. This opens up new opportunities for further development and improvement of methods for automated evaluation of acoustic quality of premises, which may have important practical implications for this field.