What do machines learn, and why? To answer these questions we import models of human cognition into machine learning …. We propose two ways of modeling machine learners based on this join: feasibility-based and cost-based machine learning …. We evaluate and estimate our models using a deep learning convolutional neural network that predicts pneumonia from chest …

We survey work using Bayesian learning in macroeconomics, highlighting common themes and new directions. First, we … present many of the common types of learning problems agents face---signal extraction problems---and trace out their effects … information. Models differ in their motives for information acquisition and the cost of information, or learning technology …

triggers large asset-price jumps. We show that the learning choice is critically determined by preference parameters and the …

We survey the recent literature on learning in financial markets. Our main theme is that many financial market … uncertain and subject to learning. We discuss phenomena related to the volatility and predictability of asset returns, stock …

Learning or experience curves are widely used to estimate cost functions in manufacturing modeling. They have recently … shows that there is a fundamental statistical identification problem in trying to separate learning from exogenous … technological change and that the estimated learning coefficient will generally be biased upwards. Second, we present two empirical …

This paper develops a theory of expectations-driven business cycles based on learning. Agents have incomplete knowledge … series is captured, unlike in standard models. Inherited from real business cycle theory, the benchmark model suffers a …

We study the perfect Bayesian equilibrium of a model of learning over a general social network. Each individual … characterize the conditions under which there will be asymptotic learning -- that is, the conditions under which, as the social … unbounded (meaning that the implied likelihood ratios are unbounded), there will be asymptotic learning as long as there is some …

a quality upgrade one of these inputs requires a period of learning before it can be used effectively, then in general … wasteful to tie up funds in the other inputs which will be underutilized until the date learning is over. We provide evidence …

, endogenous learning here eliminates that effect …

networks are recognized as the most credible source of information about new technologies. We incorporate social learning in … communicator influences learning and adoption. Farmers find communicators who face agricultural conditions and constraints most … comparable to themselves to be the most persuasive. Incorporating communication dynamics can take the influential social learning …