Introduction: The deadly disease in Indian women is Breast Cancer (BC). A mammogram is used for identifying the tumours in the breast in the early stages which is efficient and cost-effective. Objective: The main objective is to predict BC in the early stages using image processing and machine learning techniques. Methods: Our proposed methodology is 6 step process which includes preprocessing, feature extraction, feature selection, splitting the data into training and testing, classification and performance measure. Results: The experiments are done on MIAS (Mammogram Image Analysis Society) dataset. As more noise in the images of this dataset, filters are applied to get more clarity in images. Features are extracted by Local Binary Patterns (LBP) and optimized by Forest Optimization Algorithm (FOA). These features are divided into 70% training and 30% testing data for classification. The classifiers used are K- Nearest Neighbor (KNN), Na__ampersandsigniuml;ve Bayes (NB) and Random Forest (RF). Conclusion: The experiments show that LBP based FOA with RF classifier achieved good accuracy in classifying the mammograms.