Josef M. Broder

The K-armed dueling bandits problem

We study a partial-information online-learning problem where actions are restricted to noisy comparisons between pairs of strategies (also known as bandits). In contrast to conventional approaches that require the absolute reward of the chosen strategy to be quantifiable and observable, our setting assumes only that (noisy) binary feedback about the relative reward of two chosen strategies is available. This type of relative feedback is particularly appropriate in applications where absolute rewards have no natural scale or are difficult to measure (e.g., user-perceived quality of a set of retrieval results, taste of food, product attractiveness), but where pairwise comparisons are easy to make. We propose a novel regret formulation in this setting, as well as present an algorithm that achieves information-theoretically optimal regret bounds (up to a constant factor).

Dynamic Pricing Under a General Parametric Choice Model

We consider a stylized dynamic pricing model in which a monopolist prices a product to a sequence of T customers who independently make purchasing decisions based on the price offered according to a general parametric choice model. The parameters of the model are unknown to the seller, whose objective is to determine a pricing policy that minimizes the regret, which is the expected difference between the sellers revenue and the revenue of a clairvoyant seller who knows the values of the parameters in advance and always offers the revenue-maximizing price. We show that the regret of the optimal pricing policy in this model is

Determinants of Teaching Quality: What's Important to Students?.

\Theta(\sqrt T)

Monetary Returns to Bachelors and Masters Degrees in Agricultural Economics

, by establishing an

Temporal Analysis of Income Earned by Former Agriculture Students

\Omega(\sqrt T)

Estimating weak ratiometric signals in imaging data. I. Dual-channel data

lower bound on the worst-case regret under an arbitrary policy, and presenting a pricing policy based on maximum-likelihood estimation whose regret is

Time Allocation Decisions Associated with Job Market Preparation: Undergraduate Perceptions versus Market Realities

\cal{O}(\sqrt T)

Empiricism and the Art of Teaching

across all problem instances. Furthermore, we show that when the demand curves satisfy a “well-separated” condition, the T-period regret of the optimal policy is Θ(log T). Numerical experiments show that our policies perform well.

QUASI-EXPERIMENTAL DESIGNS FOR MEASURING IMPACTS OF DEVELOPMENTAL HIGHWAYS IN RURAL AREAS

A method for using student evaluations to help faculty improve their teaching performance is presented. A survey of current methods of student evaluations of teaching identified a need to improve the statistical information obtained from these evaluations. An ordinary least squares framework is used to identify the factors that students feel are important in teacher and course ratings. This framework is used to estimate weights that students assign to various teacher and course attributes and to test whether students apply these weights consistently across teachers and courses. About 81 percent of the explained variation in teacher ratings was associated with attributes that contribute to student enjoyment of the learning process. Over 90 percent of the explained variation in course ratings was associated with attributes that measure how much a student learned in the course. Students were found to apply these attributes or weights consistently across teachers and courses. Implications for developing effective teaching strategies, faculty recruitment, and curriculum reform are discussed.

Research productivity and selected characteristics of agricultural economics research and teaching faculty in the Southern Region.

Characteristics of agricultural economics alumni from the University of Georgia were discussed and a general linear model was developed to explain differences in bachelors and masters alumni salaries. Graduate education, work experience, resource mobility, gender, family background, and high school size were found to influence alumni salaries. Returns to a masters degree were computed along with capital recovery periods. Assistantship levels, cost of borrowed and human capital, and nonmonetary benefits were relevant in decisions to attend graduate school. Some costs and benefits of graduate education were found to be subjective, to vary across individuals, and to be influenced by prevailing economic conditions.