decision tree to predict the week day of the publishing of a blog post

#1

Hi
I need to predict the week day of the publishing of a blog post. the variables i will make use of in my dataset are
263…269: binary indicator features (0 or 1) for the weekday (Monday…Sunday) of the basetime
270…276: binary indicator features (0 or 1) for the weekday (Monday…Sunday) of the date of publication of the blog post

i started out with trying to combine 263-269, and also 270-276 to make the data easier to handle. i used the node column combiner. next i am trying to used several String replacers to ensure that each row only contains the day where the post was published. however i cannot configure it right and when i later use the decision tree leaner, i can’t figure it out.

Maybe there is a hole other way to solve this task still with the decision tree leaner.

0 Likes
#2

Hello @Alberte -

It’s not clear to me how your data is formatted, or what type of manipulation you’re trying to do. Perhaps you could post a sample workflow and dataset, so that we could understand your problem and approach better?

1 Like
#3

If you want to predict something with any kind of learner then my understanding is that the variable to be predicted should be in one column, not in 7.
you can try the many to one node or just make a rule (rule engine node) to create a new column with Monday_published, Tuesday_published, … as entries.

All those binary columns are sometimes just too many…we are not doing some ancient school stuff with SPSS here :wink:

1 Like