How could big data put these domains on firmer ground

  • No School
  • AA 1
  • 13

This preview shows page 10 - 12 out of 13 pages.

How could big data put these domains on firmer ground? Hastie et al enumerate that errors in prediction come from three sources 9 . The first type is from misspecification of a model. For example, a linear model that attempts to fit a nonlinear phenomenon will generate an error simply because the linear model imposes an inappropriate bias on the problem. The second source of error is from the use of samples for estimating parameters. The third is due to randomness, even when the model is perfectly specified. Figure 4: Sources of Error in Predictive Models and Their Mitigation As illustrated in Figure 4, big data allows us to significantly reduce the first two types of errors. Large amounts of data allow us to consider richer models than linear or logistic regressions simply because there is a lot more data to test such models and compute reliable error bounds. Big data also eliminate the second type of error as sample estimates become reasonable proxies for the population.
Image of page 10

Subscribe to view the full document.

The theoretical limitation of observational data, regardless of how big it is, is that it is generally “passive,” representing what actually happened in contrast to the multitude of things that could have happened had the circumstances been different. In the healthcare example, it is like having observed the use of the healthcare system passively, and now having the chance of understand it in retrospect and extract predictive patterns from it. The data do not tell us what could have happened if some other treatment had been administered to a specific patient or to an identical patient. In other words, it does not represent a clean controlled randomized experiment where the researcher is able to establish controls and measure the differential impact of treatments on matched pairs. Interestingly however, we are now in an era where there is increasing possibilities of conducting large scale randomized experiments on behavior on the Internet and uncover interesting interactions that are not possible to observe in the laboratory or through observational data alone. In a recent controlled experiment on “influence versus homophily” conducted on Facebook via an “app,” Aral and Walker uncovered the determinants of influence on online video games. 2 Their results include patterns such as “Older men are more influential than younger men,” “people of the same age group have more influence on each other than from other age groups,” etc. These results, which are undoubtedly peculiar to video games, make us wonder whether influences are different for different types of products, and more generally that influence is more complicated than we thought previously, not amenable to simple generalizations like Gladwell’s 7 concept of “super influencers.” The last of these assumptions has also been questioned by Goel et.al 8 who observe in large scale studies that influence in networks is perhaps overrated.
Image of page 11
Image of page 12

What students are saying

  • Left Quote Icon

    As a current student on this bumpy collegiate pathway, I stumbled upon Course Hero, where I can find study resources for nearly all my courses, get online help from tutors 24/7, and even share my old projects, papers, and lecture notes with other students.

    Student Picture

    Kiran Temple University Fox School of Business ‘17, Course Hero Intern

  • Left Quote Icon

    I cannot even describe how much Course Hero helped me this summer. It’s truly become something I can always rely on and help me. In the end, I was not only able to survive summer classes, but I was able to thrive thanks to Course Hero.

    Student Picture

    Dana University of Pennsylvania ‘17, Course Hero Intern

  • Left Quote Icon

    The ability to access any university’s resources through Course Hero proved invaluable in my case. I was behind on Tulane coursework and actually used UCLA’s materials to help me move forward and get everything together on time.

    Student Picture

    Jill Tulane University ‘16, Course Hero Intern

Ask Expert Tutors You can ask 0 bonus questions You can ask 0 questions (0 expire soon) You can ask 0 questions (will expire )
Answers in as fast as 15 minutes