One kind of cancer that develops in the breast’s cells is called breast cancer. In the world, it is the most prevalent cancer among people. Everyone gets breast cancer, although people are significantly more likely to have it. For those with breast cancer, early diagnosis and treatment can greatly improve prognosis and results. The study suggests using a hybrid algorithm to categorize breast cancer types. Images are initially preprocessed, which include applying the Gaussian smooth to remove noise, diminishing the image, and switching it to grayscale. PCA model is used to extract visual features. For classification, an enhanced Adam optimization ensemble including DenseNet, ResNet, XGBoost, and AdaBoost models is employed. Model performance is contrasted with those of other research findings in the literature. With an F1 score of 98.72%, accuracy of 98.8%, sensitivity of 98.15%, specificity of 98.45%, and precision of 98.44%, the suggested model has achieved impressive results.