How to built a model witrh setpoints i want

PapadPapad Computer Science StudentMember Posts: 68   Unicorn
Hi all,
I have a dataset with these attributes : 

what I want is to find correlations between speed and pressure , and make predictions about events that will happen.
I want a way to train a model, in order to can I achieve my purpose. For example, if for 10 minutes or more,pressure is above average that means that I have an event which my model has to know. How can I train my model to know it ?Which is the way in Rapidminer?
Thanks in advance.
Tagged:

Answers

  • Telcontar120Telcontar120 Moderator, RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 1,210   Unicorn
    You first need to define the attribute you are trying to predict, which in RapidMiner is called a label.  So in your example, you would need to use Generate Attributes to define something like "Pressure_Event" and then use Set Role to make it your label.  I highly recommend going through some of the built-in tutorials around model training and validation, it will make working with RapidMiner for this purpose much more clear.
    Brian T.
    Lindon Ventures 
    Data Science Consulting from Certified RapidMiner Experts
    sgenzerPapad
  • PapadPapad Computer Science Student Member Posts: 68   Unicorn
    @Telcontar120 Thanks for your answer,
    I have already seen tutorials about this, but what I want to say is when I choose a model through RM I want to give some definitions like "when pressure is under 2000 for 10 minutes or more, it means a problem in machines and you have to learn it, in order to have a prediction next time that this gonna happen".
    I hope I make it more clear.
    Thanks again.
Sign In or Register to comment.