Best Of

I agree with @varunm1 that you should try different models and optimize their hyper-parameters and indeed GBT gives you a better performance instantly. Having said this, I think that 1100 attributes … (View Post)

Hello @mansour_ebrahim Here are my observations based on your data, process and some of my earlier studies. The first thing to remember, deep learning doesn’t work all the time, this is similar to… (View Post)

Hi, see: https://github.com/rapidminer/rapidminer-studio/blob/master/src/main/java/com/rapidminer/operator/learner/functions/kernel/evosvm/PSOSVM.java feel free to commit the improvements ~Martin (View Post)

Hello @Noel Worked fine for me on a similar system. Try closing and opening again or downloading again. Sometimes these simple things can do.  @Marco_Boeck might suggest something here. (View Post)

Just came across (stole) this list from okulbilisim/awesome-datascience * Wes McKinney - Wes McKinney Blog. * Matthew Russell - Mining The Social Web. * Greg Reda - Greg Reda Personal Blog * K… (View Post)

I am not sure what you mean by "hybrid"---are you referring to ensemble models? If so, that doesn't really work for unsupervised ML problems like association rules.  Ensemble modeling requires you to… (View Post)

Hi @Joris, can you maybe try this? (Spoiler) BR, Martin (View Post)

Hi @halaalrobassy, The shortest way is to use Generate Attributes with the function cut(). The syntax is explained within the Operator (see attached screenshot). Of course you can also achieve t… (View Post)

Hi, maybe generate attributes with length(str([attribute])) ? Best, Martin (View Post)

Hello @halaalrobassy, Please find attached the process. It's pretty simple: You need to make sure your numbers are polynominal. Then with the Generate Attributes operator you can create an … (View Post)