1. 26

  2. 2

    Honest question (I know basically nothing about this area): are recommendation algorithms ever used for anything other than…selling things on the web?

    1. 5

      They only became “recommendation algorithms” recently. Before sub-specializing, they were used in information retrieval. If you recall, about 10 years ago there was the Netflix Prize to advance the field (and boy did it cause a lot of papers to be written!) by putting a million dollars out there to craft accurate prediction engines out of existing or novel new information retrieval algorithms. The team that won (submitting 24 minutes before the 3 year long deadline) was BellKor’s Pragmatic Chaos (paper here). At its heart, this super-tuned recommendation algorithm is actually an ensemble of pretty traditional IR algos: Restricted Boltzmann Machines (RBMs), matrix factorization with temporal dynamics, and a bunch of basic predictors brought together with gradient boosted decision trees (GBDT). These are general-purpose techniques that were tuned and blended to produce the recommender (RBMs can just as easily be used in computer vision and credit scoring, for example).

    2. 1

      This is way too much attention for a teen. I mean what if their idea is wrong in some small way. pop

      1. 8

        If someone of the caliber of Scott Aaronson is sufficiently convinced you are right to put his name on the paper, there is no shame in being wrong. No one in the field would hold that against Tang if it turned out to be wrong.

        I fail to see what being a teen has to do with anything. If he was a year older, or already in grad school, it would have been fine?

        To me that sounds like an argument from the same visceral response I have to these kinds of stories: jealousy. I really have to suppress the urge to rain on his parade. And doing that via arguments that seem to have someone’s best interest in mind is the socially safest way to do so. It’s a way of stealthily sabotaging someone’s accomplishment.

        1. 3

          I think what voronoipotato may have been alluding to was that aside from that age largely being turmoil for emotions as they learn to deal with them. Teenagers aren’t as battle hardened as adults and having so much attention plastered upon them for it to suddenly turn nasty can be a massive blow emotionally and without the right support can end up dissuading them from continuing.

          It sounds like the teen will be fine, with someone like Scott Aaronson on the same team - even if the idea turns out to be flawed they will have all the support they need to continue.

          1. 3

            Exactly, to be precise the problem I had was with the article pushing Tang into the limelight and not the research Tang contributed to. The research is fine.

          2. 3

            I’m not jealous at all. I don’t have any desire to get into academia or compete in that way. I stopped with a community college degree. The difference a few years can make at that age in emotional development is pretty big. Also they have other past successes however small to fall back on. Some people have one big (perceived) catastrophic failure at the beginning and they give up, and they never come back. More importantly, notoriety isn’t Tang’s accomplishment. Saying that they shouldn’t be forced into the limelight on their first attempt isn’t saying they should never have had the opportunity to submit or contribute to scientific progress.

          3. 6

            Once Tang had completed the algorithm, Aaronson wanted to be sure it was correct before releasing it publicly. “I was still nervous that once Tang put the paper online, if it’s wrong, the first big paper of [Tang’s] career would go splat,” Aaronson said.

            The article specifically mentions that point.

            1. 2

              Yeah I don’t see how that makes it any better.