When it comes to the question of female death throughout the world, breast cancer plays a significant role due to which the female mortality rate is high. So, to address this crucial question effective methods are necessary to diagnose breast cancer in an early stage to do proper treatment. Here comes the application of emerging technologies in the form of machine learning. Machine learning shows promising significance in breast cancer prediction. To address these capabilities of machine learning a thorough focused research is needed. So, in this paper, we used four machine learning algorithms named Random Forest, Random Tree, J48, and Multilayer Perceptron. We applied those algorithms to the well-known Wisconsin Diagnostic dataset on Breast Cancer. After applying feature selection techniques named Information Gain, Gain Ratio, ReliefF, and OneR we used machine learning algorithms mentioned above with 10-fold cross-validation to the given dataset. Thus, we got only 8 features which are significant out of 32 features present in the original dataset to predict breast cancer’s presence in the human body. We also try to achieve high consistency, sensitivity, and specificity levels by exploring popular ensemble approaches of algorithms in machine learning. By writing this paper we want to establish a comprehensive framework to guide breast cancer prediction using decision-making trees for the benefit of humans.