Semi-supervised multi-view learning is a remarkable but challenging task. Existing semi-supervised multi-view classification (SMVC) approaches mainly focus on performance improvement while ignoring decision reliability, which limits their deployment in safety-critical applications. Although several trusted multi-view classification methods are proposed recently, they rely on manual annotations. Therefore, this work emphasizes trusted multi-view classification learning under semi-supervised conditions. Different from existing SMVC methods, this work jointly models class probabilities and uncertainties based on evidential deep learning to formulate view-specific opinions. Moreover, unlike previous works that explore cross-view consistency in a single schema, this work proposes a multi-level consistency constraint. Specifically, we explore instance-level consistency on the view-specific representation space and category-level consistency on opinions from multiple views. Our proposed trusted graph-based contrastive loss nicely establishes the relationship between joint opinions and view-specific representations, which enables view-specific representations to enjoy a good manifold to improve classification performance. Overall, the proposed approach provides reliable and superior semi-supervised multiview classification decisions. Extensive experiments demonstrate the effectiveness, reliability and robustness of the proposed model.

Hepatocellular carcinoma (HCC) is a heterogeneous malignancy with a high incidence and poor prognosis. Exploration of the underlying mechanisms and effective prognostic indicators is conducive to clinical management and optimization of treatment. The RNA‐seq and clinical phenotype data of HCC were retrieved from The Cancer Genome Atlas (TCGA), and differential expression analysis was performed. Then, a differential lncRNA‐miRNA‐mRNA regulatory network was constructed, and the key genes were further identified and validated. By integrating this network with the online tool‐based ceRNA network, an HCC‐specific ceRNA network was obtained, and lncRNA‐miRNA‐mRNA regulatory axes were extracted. RNAs associated with prognosis were further obtained, and multivariate Cox regression models were established to identify the prognostic signature and nomogram. As a result, 198 DElncRNAs, 120 DEmiRNAs, and 2827 DEmRNAs were identified, and 30 key genes identified from the differential network were enriched in four cancer‐related pathways. Four HCC‐specific lncRNA‐miRNA‐mRNA regulatory axes were extracted, and SNHG11, CRNDE, MYLK‐AS1, E2F3, and CHEK1 were found to be related with HCC prognosis. Multivariate Cox regression analysis identified a prognostic signature, comprised of CRNDE, MYLK‐AS1, and CHEK1, for overall survival (OS) of HCC. A nomogram comprising the prognostic signature and pathological stage was established and showed some net clinical benefits. The AUC of the prognostic signature and nomogram for 1‐year, 3‐year, and 5‐year survival was 0.777 (0.657‐0.865), 0.722 (0.640‐0.848), and 0.630 (0.528‐0.823), and 0.751 (0.664‐0.870), 0.773 (0.707‐0.849), and 0.734 (0.638‐0.845), respectively. These results provided clues for the study of potential biomarkers and therapeutic targets for HCC. In addition, the obtained 30 key genes and 4 regulatory axes might also help elucidate the underlying mechanism of HCC.