In-order to learn MapReduce algorithm,I usually try it from scratch. What I follow is for classification algorithms-
*First I build a model * *hadoop jar myjar.jar edu.ModelDriver <Modelinput> <Modeloutput>* *secondly I will write a prediction class in MR* *hadoop jar myjar.jar edu.PredictDriver <Testinput> <TestOutput> <Modeloutput>* *<Modeloutput> is supplied as an argument to get the model results for prediction.* Is this a better way ? or Should I follow the below way - *hadoop jar myjar.jar edu.Driver train="traininput.txt" test="testinput.txt" output=outputlocation* Which is the standard way? Please suggest -- *Thanks & Regards * *Unmesha Sreeveni U.B* *Hadoop, Bigdata Developer* *Center for Cyber Security | Amrita Vishwa Vidyapeetham* http://www.unmeshasreeveni.blogspot.in/
