1. Apply an attribute selection method on the project-training.csv (or project-training arff ) dataset to select a subset of attributes. Let's call this reduced training dataset . From the project- test.csv (or project-test.arff) dataset, select only those attributes that are in the reduced training dataset. Let's call this reduced test dataset . So, the reduced training dataset and the reduced test dataset have exactly the same attributes. 2. Build a classifier model using a classification algorithm from the reduced training dataset and test it on the reduced test dataset. Collect the test result. Repeat this four more times with four different classification algorithms. So, you will build and test five classifier models with the same reduced training dataset and the same reduced test dataset