Author(s): Lirong Xia, Ronald Brachman, Francesca Rossi, Peter Stone
The ubiquitous problem of studying and decision-making from rank information arises in eventualities the place clever programs acquire choice and behaviour information from people, be told from the knowledge, after which use the knowledge to assist people make environment friendly, efficient, and well timed choices. Often, such information are represented via scores.
This e-book surveys some contemporary development towards addressing the problem from the issues of statistics, computation, and socio-economics. We will duvet classical statistical fashions for rank information, together with random software fashions, distance-based fashions, and aggregate fashions. We will talk about and examine classical and state of-the-art algorithms, equivalent to algorithms in accordance with Minorize-Majorization (MM), Expectation-Maximization (EM), Generalized Method-of-Moments (GMM), rank breaking, and tensor decomposition. We may also introduce principled Bayesian choice elicitation frameworks for accumulating rank information. Finally, we will be able to read about socio-economic sides of statistically fascinating decision-making mechanisms, equivalent to Bayesian estimators.
This e-book can also be helpful in 3 ways: (1) for theoreticians in statistics and device studying to raised perceive the issues and caveats of studying from rank information, in comparison to studying from different varieties of information, particularly cardinal information (2) for practitioners to use algorithms lined via the e-book for sampling, studying, and aggregation and (3) as a textbook for graduate scholars or complex undergraduate scholars to be told concerning the box.
This e-book calls for that the reader has elementary wisdom in likelihood, statistics, and algorithms. Knowledge in social selection would additionally assist however isn’t required.