Aggregating performance measures considerably reduces the complexity of a performance evaluation system. This dissertation analyzes the nature and characteristics of the aggregation of performance measures in institutional settings of multiple tasks and multiple periods. In multi-task settings, the number of tasks restricts the feasibility of a statistically sufficient aggregation and the nature of an economically sufficient aggregation. Statistical sufficiency of aggregation can be achieved only by multi-dimensional aggregate measures, whose minimum dimensionality is given by the number of tasks. If the number of aggregate measures is less than the number of tasks, an economically sufficient aggregation incurs loss of information even if there is no loss of information through the likelihood ratio. The results support the use of multi-dimensional aggregate measures to preserve the information content of performance measures in multi-task agencies. In a multi-period setting, the inter-temporal correlations among performance measures restrict the feasibility of a statistically sufficient aggregation and the nature of an economically sufficient aggregation. When performance measures are inter-temporally correlated, there is no statistically sufficient aggregation and an economically sufficient aggregation of the basic measures depends on the effort level to induce. The optimal aggregation is characterized by the agent's characteristics and the economic situation of the agency as well as the statistical properties of performance measures. In a long-term contract with multiple tasks, the inter-temporal covariance risk has a monotonic impact on the endogenous allocation of effort through the optimal relative incentive rate. The inter-temporal covariance risk, as well as the within-period risk premium, prevents the first best allocation of effort from being endogenously achieved even if the first best allocation is feasible. / Accounting

Increasingly, predictive models are used to support routine business de- cisions and are integral to the strategic competitive business strategies for a wide range of industries. Most often, data-driven predictive models are in- duced from training data obtained through the businesss routine operations. However, recent research on policies for intelligent information acquisitions suggests that proactive acquisition of information can improve models at a lower cost. Most active information acquisition policies are accuracy centric; they aim to identify acquisitions of training data that are particularly benefi- cial for improving the predictive accuracy of a given model. In practice, however, inferences from a predictive model are often used along with inferences from other predictive models as well as constant factors to inform arbitrarily complex decisions. In this dissertation, I discuss how these settings motivate a new kind of collaborative information acquisition (CIA) policies that exploit knowledge of the decision to allow multiple predictive models to collaboratively prioritize the prospective information acquisitions, so as to best improve the decisions they inform jointly. I present a framework for CIA policies and two specific CIA policies: CIA for binary decisions (CIA-BD), and CIA for top-ranked opportu- nities in terms of expected revenue (CIA-TR). Extensive empirical evaluations of the policies on real-world data suggest that the notion of CIA policies is indeed a valuable one. In particular, I demonstrate that these two new poli- cies lead to superior decision-making performances as compared to those of alternative policies that are either decision-centric or do not allow multiple models to collaboratively prioritize acquisitions. The performance exhibited by the CIA policies suggest that these policies are able to effectively exploit knowledge of the decisions to avoid greedy improvements in accuracy of any individual model informing the decisions; instead, they promote improvements in any one or all of the models when such improvements are likely to benefit the decisions. / text

No description available.

International Telemetering Conference Proceedings / October 28-31, 1996 / Town and Country Hotel and Convention Center, San Diego, California / This paper presents a new architecture of the end-to-end ground system to reduce overall mission support costs. The present ground system of the Jet Propulsion Laboratory (JPL) is costly to operate, maintain, deploy, reproduce, and document. In the present climate of shrinking NASA budgets, this proposed architecture takes on added importance as it will dramatically reduce all of the above costs. Currently, the ground support functions (i.e., receiver, tracking, ranging, telemetry, command, monitor and control) are distributed among several subsystems that are housed in individual rack-mounted chassis. These subsystems can be integrated into one portable laptop system using established MultiChip Module (MCM) packaging technology. The large scale integration of subsystems into a small portable system will greatly reduce operations, maintenance and reproduction costs. Several of the subsystems can be implemented using Commercial Off-The-Shelf (COTS) products further decreasing non-recurring engineering costs. The inherent portability of the system will open up new ways for using the ground system at the “point-of-use” site as opposed to maintaining several large centralized stations. This eliminates the propagation delay of the data to the Principal Investigator (PI), enabling the capture of data in real-time and performing multiple tasks concurrently from any location in the world. Sample applications are to use the portable ground system in remote areas or mobile vessels for real-time correlation of satellite data with earth-bound instruments; thus, allowing near real-time feedback and control of scientific instruments. This end-to-end portable ground system will undoubtedly create opportunities for better scientific observation and data acquisition.