Rapidminer Tutorials

Welcome to Neural Market Trend's "Tutorials" lesson page. Here you can learn how to build predictive trend model using neural nets and artificial intelligence (AI), use Genetic Algorithms, and build analytic trend systems in Excel. Be sure to visit my entire set of lessons by selecting the tutorial category! Rapidminer Sample Processes Rapidminer 5.0 Video Tutorials

14 Comments

Hi Thomas,
Where can I download your original written tutorial you mention in your videos?

See this page for links to the written YALE/Rapidmner tutorials blog posts. The original ones were in written form but are very outdated now because RM has gone under heavy changes.

Hi guy. Your video-tutorials are excellent. I downloaded those about rapidminer to keep as reference.
I am trying to use it to predict stock market too. Good job!

Thanks! Good luck!

These videos are very interesting and useful.
Thank's Thomas, keep going.

Thanks, I'll try.

thanks a lot mr. thomas ,.. Your tutorial is really great and helpfull.

Hello Thomas,
Your tutorials are really wonderful. I tried using them and they were really helpful.
I have a small problem at the moment for which I am seeking solution. Any help would be appreciated.
1) Which would be the best algorithm/operator in rapid miner that could help predict 5 outputs from a set of inputs.
I have tried using neural nets but id doesnt give me an accuracy as expected. Now I am playing with evolutionary algorithms to help select best inputs for my neural network.
2) My inputs are not numbers. They are mainly character data type. How can evolutionary algorithm be used to weigh these type of inputs based on a status attribute (which unlike the gold status in your example has 5 values namely active, terminated, withdraw, pause, dropout)
Regards

@Vaishak: I don't know enough about your data to make a definitive answer but I suggest you check out using a SVM operator. However, you'd have to transform your inputs to numerical attributes, easy to do with the Nominal to Numerical operator, and make sure your outputs are also numerical labels. Make sure you use a Radial kernel when using the SVM operator.

@Awatif: I'd love too but I have no time right now. I'm overloaded with work.

Dear Mr Tom,
Once again thank you for that wonderful video #13 on parameter optimization using Neural Nets. It was really helpful.
I tried the same using SVM but I guess I have some memory issues.
Iam new to SVM and its parameters seem to make no sense. Though I changed a few of them like C, gamma, Kernal types refering your previous videos, it would have been good to know
1) What these parameters are
2) How each could infuluence the inputs.

I went through some research materials but I guess I am finding it hard to understand them. Is there any advice or help you could offer.
Regards
Vaishak

Hi Tom,

First and foremost thank you for posting these videos. They are a great way to get to know the tool.

I am commenting to say that the following link (video 10) is broken:
http://www.neuralmarkettrends.com/2010/04/rapidminer-5-0-video-tutorial-10-%e2%80%93-financial-time-series-modeling-part-2.html

I am using Linux/Chrome but don't think that is the issue. The problem are the characters after the video number (10-%e2%80%93). If I use double "-", it works correctly (10--) for example:

http://www.neuralmarkettrends.com/2010/04/rapidminer-5-0-video-tutorial-10--financial-time-series-modeling-part-2.html

Video 11's link also seems to be broken.

Regards,
Hugo F.

Hi Tom

I assume the pattern recognition plugin only works for classification problems, is this correct? Are there automatic tools for non-classification problems, like regression? Also what operators does Rapid Miner have for regression problems?

Thanks

Larry

hi tom,
It was my first encounter with rapid miner and with ur video,both look awesome,thnx for such a easy to understand and knowledgeable video.U have really got me interested in this.Thanks a lot.

Leave a comment

About this Archive

Find recent content on the main index or look in the archives to find all content.