Clinical validation of lightweight CNN architectures for reliable multi-class classification of lung cancer using histopathological imaging techniques.

The results showed a clear trend: as the network depth increased, the best validation F1 score improved, i.e., \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$F_1^{\text {Lite-V0}}< F_1^{\text {Lite-V1}}< F_1^{\text {Lite-V2}} < F_1^{\text {Lite-V4}},$$\end{document} indicating that deeper architectures capture more discriminative features in histopathological images.

also in 19,460 other papers