Convergence of CNN and SVM Techniques for Liver Cancer Classification: Assessing Molecular, Clinical Presentation, and Staging Classifications

The field of the research is introduced briefly together with the methods and general findings of the study. Three primary areas of classification that are molecular classification, clinical presentation classification, & staging classification are applied in a meticulous manner to assess the efficacy of this classification model. Based on the analysis of the results, it is evident that the proposed model surpasses all the other models in the molecular classification achieving an accuracy, recall, as well as F1-score of 95% at most. It also showed a 89% function of detecting and categorising the molecular features of liver cancer. The model also works well in giving high accuracy in distinguishing liver cancer cases according to clinical manifestations and disease progression, seen by the very high precision, recall & F1-score of 0.95. 24 % & 96. The identified disparities were beneficial with patients in the amino acid pattern achieving a 17% increase in the presentation of clinical symptoms and a 30% improvement in the staging classification. The model was trained on the dataset with 10,850 photos, and the accuracy of the work was 95%. 85%. The macro, weighted & micro average metrics are also mentioned in the abstract which guarantees the usage of the model for various classification. On it, the following aspects are also described: support & accuracy; these are the indicators that show the support the model gives to particular classes, along with the total accuracy of the results. All these comprehensive evaluations illustrate the performance of the introduced model in the classification of liver cancer with high accuracy and low variability, enriching the understanding of the disease and its treatment.