reproduction of results

Dear Community,
I will reproduce this result in the following picture.
With this process.
I have the right data set to reproduce this result but it doesnt work. The problem is that I don't have the settings of the process items (select attribute...).
What should I change in the process to get the same results with this process like in picture one?
Thank you
Answers
-
Those results are a combination of the parameters you choose in the FP-Growth and Create Asssociation Rules operator. Did you try adjusting those?
0 -
Dear Thomas,
Thank you for your answer.
I didn't try this. How is ist possible to combine parameters in fp growth?
0 -
You click on each operator and then set the parameters that are located (usually) on the right side of Studio.
0 -
-
I will describe my problem exactly.
I have this set of data with this set of data https://1drv.ms/x/s!AguC3J6JV9sYvUuhvFYzY7wIMuPJ
My question is, what charaktarice a person who buys a lot of beer?
And the other question is, what charaktarice a person who buys a lot of crisps?
And the result must be like in picutre one.
Who must look my association analyses in rapid miner to get the result like the picture with this given set of data?
Thank you for your answers
0 -
Is it possible to show a screen shot with this parameter option?
0 -
This is the standard beer and diaper market basket analysis. If you can't see the parameter panel, then you probably closed it by mistake. Go to View > Restore Default and then click on the FP-Growth operator. You should see this.
0 -
Thank you for this screenshot.
But I cant combine here some parameters.
How would you like to design the whole process to get the same table like in the picture ?
regards,
Phares
0 -
Did you create the process you screenshot above? You will need to adjust the parameters to get the right confidence level, etc. that will match the output. Is this some university homework?
0 -
Did you create the process you screenshot above? You will need to adjust the parameters to get the right confidence level, etc. that will match the output. Is this some university homework?
0 -
-
I dunno, this is where you need to experiment.
That said, you should probably do some ETL work on the data. There's a lot of data that could be transformed to make this process faster
0