Skip to main content
added 7 characters in body
Source Link

Let an m by n matrix represent a set of movie-person preference data; n columns representing n different movies, m rows representing m different persons, and each entry shows a particular person's preference value of a particular movie.

Column space: a set of different movies with unique fandom, that is preference profile. Row space: a world of movie-viewers, each person with different (possibly independent) taste.

Then the null space (kernal) can be thought of as a possible movie that is nat (and can not be) liked by any person. The left null space would be an essentric person with a weird taste that cannot be satisfied any existing movie or any future movies that can be made independently from the existing movies. Those movies and persons, however, do exist in the entire movie-person space.

Let an m by n matrix represent a set of movie-person preference data; n columns representing n different movies, m rows representing m different persons, and each entry shows a particular person's preference value of a particular movie.

Column space: a set of different movies with unique fandom, that is preference profile. Row space: a world of movie-viewers, each person with different (possibly independent) taste.

Then the null space (kernal) can be thought of a possible movie that is (and can not be) liked by any person. The left null space would be an essentric person with a weird taste that cannot be satisfied any existing movie or any future movies that can be made independently from the existing movies. Those movies and persons, however, do exist in the entire movie-person space.

Let an m by n matrix represent a set of movie-person preference data; n columns representing n different movies, m rows representing m different persons, and each entry shows a particular person's preference value of a particular movie.

Column space: a set of different movies with unique fandom, that is preference profile. Row space: a world of movie-viewers, each person with different (possibly independent) taste.

Then the null space (kernal) can be thought of as a possible movie that is nat (and can not be) liked by any person. The left null space would be an essentric person with a weird taste that cannot be satisfied any existing movie or any future movies that can be made independently from the existing movies. Those movies and persons, however, do exist in the entire movie-person space.

Source Link

Let an m by n matrix represent a set of movie-person preference data; n columns representing n different movies, m rows representing m different persons, and each entry shows a particular person's preference value of a particular movie.

Column space: a set of different movies with unique fandom, that is preference profile. Row space: a world of movie-viewers, each person with different (possibly independent) taste.

Then the null space (kernal) can be thought of a possible movie that is (and can not be) liked by any person. The left null space would be an essentric person with a weird taste that cannot be satisfied any existing movie or any future movies that can be made independently from the existing movies. Those movies and persons, however, do exist in the entire movie-person space.