Assessing the Efficacy of CNN-SVM Hybrids in Breast Tumor Recognition

Advanced diagnostic techniques are required to ensure that breast cancer is diagnosed right from the time it begins to manifest and is one of the leading causes of illness and death among women worldwide. The work in this research describes a dual model of CNN and SVM to improve the detection of breast tumors. The model underwent evaluation utilizing datasets from five distinct customers, each consisting of photos classified into five categories: benign is used to refer to conditions that are not dangerous, malignant is used to refer to cancerous conditions, cystic is an abnormal sac-like structure, and fibroadenoma is an abnormal breast tissue growth and finally normal refers to a healthy status. The hybrid model yielded the best results with an average accuracy of 95 percent, with most analyzed parameters giving satisfactory performance results. 64%, and the Gig Economy software solution accuracy of 98.4%, and the accuracy of the feature selection model is 86—$6 \%$. The hybrid model caused enhancements in the precision and recall ratio, as demonstrated in the different charts throughout the manuscript. In particular, it showed 94 percent accuracy in predictions related to newscasts. 8% Their acceptability level stands at 8%, while the recall level of the brand was $94 \% .5 \%$. The former CNN model obtained a higher precision of about $91: 8 \% ~\&$ a recall of 91%. The first model achieved an accuracy of 2%, while the other one, the SVM model, achieved a precision of 87.9%, While the precision for the second search strategy was 83.5%. The tests for client-model correspondence confirmed the model’s stability during the analysis, which resulted in an accuracy rate of 94.5% to 95. ${7 \%}$ across several datasets.