-1

I'm noob in machine learning, but I'm trying to know more about it. I have a question about a prediction model (predict for every page when the number of click) how can I do with no data in some columns.

Should I drop these rows?

Any idea please? Thanks.

Firebug
  • 15,262
  • 5
  • 60
  • 127
user17241
  • 129
  • 1
  • 1
  • 7

1 Answers1

2

If you mean that some rows have missing data (missing attributes) you can:

  • a) Drop them (if they are no too many rows with missing attributes).

If you can't / don't want to drop them then:

  • b) Fill them using K-nearest neighbors.

  • c) Use some method that can deal with missing data.

Take a look, for instance, at this thread.

Community
  • 1
alberto
  • 2,646
  • 16
  • 36
  • You're welcome @user17241. However, now the question is quite different and became a question about R, you should open a new question for that instead of changing the sense of the original that much. I guess the problem is that hotelX has not the right shape. Add the 'R' tag or even ask it in stack overflow. You might also show us the dimensions of hotelX, dim(hotelX) – alberto Feb 01 '15 at 10:24