The objective of this assignment is to be familiar with the KNIME platform and specifically to practice training a binary classification model using KNIME. Review the KNIME lab we did in class on binary classification with Decision Trees algorithm and train a decision-tree based classifier yourself, on a new binary classification problem of your choice.
Choose a dataset of your choice that is suitable for binary classification (i.e., the outcome variable is categorical in nature and has two classes). It is good if you can work with a file with no missing values.
UCI Machine Learning Repository and Kaggle, maybe good places to search for free datasets.
Alternatively, you can use Heart.csv (that I share with you) to predict heart disease, if you like. In this dataset, the “AHD” variable indicates whether the given patient has heart disease.
Use KNIME on your dataset to train a decision-tree classifier to predict the outcome. Cover all aspects we did in class that includes, data reading, any transformations needed, data partitioning, training, testing, and reporting evaluation metrics.
Prepare a document including the following:
Briefly describe the chosen dataset. Include the link to download it.
Clearly state the purpose of your model. What are you trying to predict?
Include a picture of your workflow (you can take a screenshot).
Include a picture of the decision tree you created (you can export)
Report the model performance results (confusion matrix, accuracy, precision, recall, F1, etc.) Comment on the reliability of your model.