miRNAs are small noncoding RNA molecules, mainly responsible for post-transcriptional control of gene expressions. Machine learning is becoming more and more widely used in breast tumor classification and diagnosis. In this paper, we compared the performance of different machine learning methods, such as Random Forest (RF), eXtreme Gradient Boosting(XGBoost) and Light Gradient Boosting Machine(LightGBM), for miRNAs identification in breast cancer patients. The performance comparison of each algorithm was evaluated based on the accuracy and logistic loss and where LightGBM was found better performing in several aspects. hsa-mir-139 was found as an important target for the breast cancer classification. As a powerful tool, LightGBM can be used to identify and classify miRNA target in breast cancer.