Recent collaborative applications and enterprises very often need to efficiently integrate their access control policies. An important step in policy integration is to analyze the similarity of policies. Existing approaches to policy similarity analysis are mainly based on logical reasoning and Boolean function comparison. Such approaches are computationally expensive and do not scale well for ...

As more research studies incorporate next-generation sequencing (including whole-genome or whole-exome sequencing), investigators and institutional review boards face difficult questions regarding which genomic results to return to research participants and how. An American College of Medical Genetics and Genomics 2013 policy paper suggesting that pathogenic mutations in 56 specified genes shou...

The Australian Red Cross Blood Service (ARCBS) recently introduced a policy of compulsory appointments for blood donations. This thesis examines the effect of these appointments on donor satisfaction and donation behaviour. Overall, aggregate tests indicate that the policy transition initially had a negative effect. However, conditional on having donated once after the transition, donors are mo...

This study searched the available needle exchange program (NEP) literature for return rate data. A total of 26 articles were found. The overall worldwide return rate was 90%, although this ranged from a low of 15% to a high of 112%. U.S. NEP return rates were gathered from only eight studies, indicating a clear need for more data, although U.S. return rates were comparable to those from NEPs ou...

In the field of reinforcement learning there has been recent progress towards safety and high-confidence bounds on policy performance. However, to our knowledge, no methods exist for determining high-confidence safety bounds for a given evaluation policy in the inverse reinforcement learning setting—where the true reward function is unknown and only samples of expert behavior are given. We prop...

This technical report describes a cute idea of how to create new policy search approaches. It directly relates to the Natural Actor-Critic methods but allows the derivation of one shot solutions. Future work may include the application to interesting problems. 1 Problem Statement In reinforcement learning, we have an agent which is in a state s and draws actions a from a policy π. Upon an actio...

The remarkable appreciation of the U.S. dollar after 1980 has been viewed by many observers as a failure of the floating exchange rate system, and has been a major stimulus to calls for a return to a more managed global exchange rate system. Critics of the current international monetary arrangements argue that tighter international "rules of the game" in macroeconomic policymaking would reduce ...

Reinforcement learning algorithms are typically geared towards optimizing the expected return of an agent. However, in many practical applications, low variance is desired to ensure reliability algorithm. In this paper, we propose on-policy and off-policy actor-critic that optimize a performance criterion involving both mean return. Previous work uses second moment estimate indirectly. Instead,...

The underlying rationale for this review is that policy implications cannot be automatically derived from analyses of innovation issues i.e. ‘normative’ does not automatically follow from ‘positive’. A view that policy application can be developed as an afterthought of positive analysis is similar to the logic of a linear innovation model whereby normative is derived in straightforward manner f...

We extend the set of regulatory instruments for banks’ liquidity provision by adding a policy instrument for controlling the fraction of perfectly-liquid accounts. We demonstrate how this instrument induces self-selection on behalf of depositors who are differentiated according to their probability of facing a liquidity shock. This selfselection leads to a market segmentation, which can break t...