Hi all, Anyone have any experience with this particular software? Briefly, I'm currently speaking with some programmers to help me on a machine learning project, but considering their rates and pricing, I was wondering if I might as well go with RapidMiner or something similar. Thanks in advance.
I used RapidMiner about a decade ago. I thought it was pretty good, but not sure you need all the ML algorithms that it has. My take on optimization is that it tends to lead to curve fitting -- a system that does really great on past data, but not so great on new data. Through cross-validation and other methods, there are ways to guard against that. Anyhow, I think RapidMiner is a good place to start. You can always develop something more complicated / customized later if needed.
What are you trying to accomplish with your ML project? I might be able to help. RapidMiner seems interesting but it looks like it limits processing based on your license. The solution I offer has no such limitations. Might be a good fit for you....maybe not. Whatever the case...good luck on the project!
Without going into too much detail, let's just say that I have sets of data that I'd like to explore further and I've been curious if ML could be helpful. Just something I'm interested in exploring at this point.
Sounds good. Whatever solution you go with, make sure it there will be a high probability of success before shelling out any money.
Well, that's always hard to know in advance even though RapidMiner offers a free trial. Time is the most valuable resource currently. Not in an immediate rush on this. You mentioned having a solution...?
Also consider KNIME. https://www.knime.com/ https://www.youtube.com/user/KNIMETV I've tried both years ago.
It was pretty easy to pick up if you have some basic knowledge of the ML algorithms that you are using. I also used KNIME as well, but I don't recall the difference between the two. I fed in a bunch of technical indicators and tried to predict short-term price direction. It was a misguided project and mostly a waste of time. I think trying to find correlations between different products (S&P vs. oil vs. gold, etc) might be a better use of time, but never tried it.