How can using randomly delete method?

StanleyTasStanleyTas MemberPosts:6Newbie
inHelp
This is my process.



and this is my result



but, this result is error.

I need to show this one↓



who can tell me, I lost what's attribute?

this is csv file.


spiral.csv 11.3K

Answers

  • MartinLiebigMartinLiebig Administrator, Moderator, Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert, University ProfessorPosts:3,389RM Data Scientist
    I am sorry, but from your description it is unclear what your problem is?

    Best,
    Martin
    - Head of Data Science Services at RapidMiner -
    Dortmund, Germany
  • StanleyTasStanleyTas MemberPosts:6Newbie
    edited August 2020
    I mean, how can use csv file to display the result.

    Task:Randomly delete 70% of the points of the outer spiral (see figures below).



    我尝试过更多的方法,但我ways fail

  • Telcontar120Telcontar120 Moderator, RapidMiner Certified Analyst, RapidMiner Certified Expert, MemberPosts:1,635Unicorn
    I think you probably need to define two clusters using a method like DBSCAN, which will isolate the two spirals from each other. Once you have that, then you can generate a random number for the first cluster (or the one that corresponds to the outer spiral) and then remove 70% of those based on a simple numerical threshold.
    Brian T.
    Lindon Ventures
    Data Science Consulting from Certified RapidMiner Experts
  • StanleyTasStanleyTas MemberPosts:6Newbie
    I have to use DBSCAN operation, but I don't know why cannot succeed.
  • StanleyTasStanleyTas MemberPosts:6Newbie
    who knows how can solve it?
  • MartinLiebigMartinLiebig Administrator, Moderator, Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert, University ProfessorPosts:3,389RM Data Scientist
    Hi,
    what you probably want to do: Apply DBScan on your data, check for the cluster which makes the outer ring, Filter for the outer ring, sample it down.

    This should be enough to solve this. We are happy to help here with questions to our software. We are however not doing your complete assignments.

    Best,
    Martin
    - Head of Data Science Services at RapidMiner -
    Dortmund, Germany
  • StanleyTasStanleyTas MemberPosts:6Newbie
    that's no assignment, it just exercise:)
  • MartinLiebigMartinLiebig Administrator, Moderator, Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert, University ProfessorPosts:3,389RM Data Scientist
    Than I am sorry, pretty much looks like an assignment from a university.
    - Head of Data Science Services at RapidMiner -
    Dortmund, Germany
Sign InorRegisterto comment.