A Hybrid Deep Learning Model for Apartment and Industrial Building Classification

This research evaluates the performance of a classification model in distinguishing between two types of buildings: apartments and industrial structures. Using a dataset of 10,440 images, the model demonstrated high accuracy and precision in classification. The precision for the apartment class was 98.06%, recall was 99.02%, and the F 1 -score was 98.54%.On the contrary, for the industrial class, it was 99.05%, the recall was at $\mathbf{9 8. 1 2 \%}$, and an $\mathbf{F} 1$-score of $\mathbf{9 8. 5 9 \%}$. It would indicate the minimal error by the model while distinguishing between both kinds of buildings. The test set had $\mathbf{5 1 1 0}$ apartment images and 5330 industrial images, which means it accounted for 49% and 51% of the entire data set. Despite such a small imbalance, the model got an overall accuracy of 99% for apartment as well as industrial classes, showing that it was wellbalanced and quite efficient in differentiating between these two types of buildings without any bias. Therefore, the overall accuracy of the model is an impressive 98.56%, which indicates good generalization over unseen data. The results indicate that the model is quite robust and reliable, and hence it may be applicable in real-world scenarios in urban planning, building type identification, and other related fields. It reflects the quality of the proposed model regarding such huge datasets, and how it would classify both the apartments and industrial buildings based on a good recall, precision, and F1-score for both classes and overall good accuracy.