1

Can anyone please tell me How to normalize rating in a scale of 1 to 5?

In Yahoo! Movies dataset user has given a rating to a movie on the scale of 1 to 13 and 0 means there is no rating to that movie.

Please tell me how to normalize it.. original rating in scale of 1 to 13 the new rating that needed in the program is 1 to 5.

shweta
  • 13
  • 1
  • 3
  • I think your question is already answered [here](https://stats.stackexchange.com/questions/70801/how-to-normalize-data-to-0-1-range?rq=1). – bizzinho Jun 20 '17 at 06:59

1 Answers1

2

You can use the standard re-scaling formula, i.e. $value_{new} = \frac{max_{new} - min_{new}}{max_{old} - min_{old}}\times (value_{old} - max_{old}) + max_{new}$.

In your case, that would be $\frac{5-1}{13-1} \times (value_{old} - 13) + 5$. And $value_{old} = 0 = value_{new}$.

Yannis Vassiliadis
  • 358
  • 1
  • 8
  • 1
    Usually, "a scale of 1 to 5" means the values are *integral*. Your solution does not produce integral values. The obvious solution is to round the results, but there are difficulties. The most problematic issue is that since $13/5$ is not integral, the number of original ratings that map to a new rating will vary between $2$ and $3$, thereby potentially creating some bias. That needs to be understood and, perhaps, controlled in some way. – whuber Jun 20 '17 at 13:50