I've been using KNIME for couple of days now so I'm still a rookie.
My idea is as follows - I'd like to create a flow using the classification trees to analyse clients' churn.
1. All the data is in xls file and the outcome is produced by XLS Writer. Say, I need to check someone's inclination to churn. I put new data in xls file and I'd like to get my feedback in the second xls file (created by Writer). Do I have to access KNIME and refresh entire flow every time manually or is there a way to go about it? (data is inserted and the outcome based on it appears in the second xls sheet manually with no need to re-execute)
2. Is it a good idea to divide clients' population into clusters and then divide it again using regression trees? Is the outcome understandable ?
3. (WEKA) What is the proper output for a random forest node? I've checked literally everything.
Thank you very much. Hopefully everything becomes clearer with time !