First provide a summary of the paper, and then address the following criteria: Quality, clarity, originality and significance. This paper presents a rational model of economic choice, where values are replaced by inferences about relative rank in a context. The model is shown to provide a good fit to pricing data and risky-choice experiments.Overall, I found the paper interesting, but somewhat too narrow and inaccessible for a NIPS audience. There was lots of insufficiently explained economic jargon (e.g. Veblen and Giffen goods), which are key for understanding the contribution.

First provide a summary of the paper, and then address the following criteria: Quality, clarity, originality and significance. The paper introduces a GP-Vol model to flexibly capture the time-dependent changes in variance, and develops a new online algorithm for fully Bayesian inference under the model. The paper is clearly written, the developed inference method seems technically sound, and the presented results look promising. My opinion on the model itself, using a non-parametric approach such as using the GP prior on the transition function (as in the paper), seems, though, a bit an obvious way of extending the prior work developed in the finance area. So, I wouldn't put too high grade on the paper in terms of its originality.

First provide a summary of the paper, and then address the following criteria: Quality, clarity, originality and significance. This paper presents a new Bayesian nonparametric model for mixed-membership modeling of grouped data. The model is based on a Beta-Negative Binomial Process (BNBP). Whereas this stochastic process was exploited by previous work, as mentioned in the paper, there are several aspects that set this work apart. First, this work obtains an EPPF (exchangeable partition probability function), and derives the prediction rules thereon.

So, most of the learning happens under model misspeci-fication.

In simple perceptual decisions the brain has to identify a stimulus based on noisy sensory samples from the stimulus. Basic statistical considerations state that the reliability of the stimulus information, i.e., the amount of noise in the samples, should be taken into account when the decision is made. However, for perceptual decision making experiments it has been questioned whether the brain indeed uses the reliability for making decisions when confronted with unpredictable changes in stimulus reliability. We here show that even the basic drift diffusion model, which has frequently been used to explain experimental findings in perceptual decision making, implicitly relies on estimates of stimulus reliability. We then show that only those variants of the drift diffusion model which allow stimulus-specific reliabilities are consistent with neurophysiological findings. Our analysis suggests that the brain estimates the reliability of the stimulus on a short time scale of at most a few hundred milliseconds.

Expectation propagation (EP) is a deterministic approximation algorithm that is often used to perform approximate Bayesian parameter learning. EP approximates the full intractable posterior distribution through a set of local approximations that are iteratively refined for each datapoint. EP can offer analytic and computational advantages over other approximations, such as V ariational Inference (VI), and is the method of choice for a number of models. The local nature of EP appears to make it an ideal candidate for performing Bayesian learning on large models in large-scale dataset settings. However, EP has a crucial limitation in this context: the number of approximating factors needs to increase with the number of data-points, N, which often entails a prohibitively large memory overhead. This paper presents an extension to EP, called stochastic expectation propagation (SEP), that maintains a global posterior approximation (like VI) but updates it in a local way (like EP). Experiments on a number of canonical learning problems using synthetic and real-world datasets indicate that SEP performs almost as well as full EP, but reduces the memory consumption by a factor of N . SEP is therefore ideally suited to performing approximate Bayesian learning in the large model, large dataset setting.

This approach yields biased estimates of the expected risk, and therefore leads to poor decisions for two reasons.

Neural Information Processing Systems http://nips.cc/