Lung cancer is caused by the formation of a significant number of aberrant cells, which can lead to death rates in some situations. So it is really important to predict lung cancer in the early stage which helps in saving lives. Early prediction of cancer helps to determine the level of diagnosis at which the patient should be treated, which helps doctors, the medical team, and saves money. In this paper we have proposed a framework for the early prediction of lung cancer using various machine learning algorithms. This study employs an approach for which feature selection techniques are used in combination with a classification model to predict lung cancer. The given data set is preprocessed. Synthetic Minority Oversampling Technique(SMOTE) is used to balance the data set. Outliers are used to provide insight. In this study, we employed six machine learning (ML) algorithms: KNN (k-nearest neighbour), Decision Tree (DT), Random Forest (RF), XGBoost, Naive Bayes (NB), and Logistic regression (LR) models. Principal Component Analysis(PCA) is used for feature selection along with low variance filter which helps in selecting the features which show a major impact on the output. Pipelines are used to streamline the process and helps in easier modification of steps. Cross-validation is used to increase the reliability. The model’s performance is assessed using performance measures. After a rigorous experimentation, performance of the different models are presented with a extensive analysis.