This paper describes and evaluates the application of three local learning algorithms --- nearest-neighbor, weighted-average, and locally-weighted polynomial regression --- for the prediction of run-specific resourceusage on the basis of run-time input parameters supplied to tools. A two-level knowledge base allows the learning algorithms to track short-term fluctuations in the performance of computing systems, and the use of instance editing techniques improves the scalability of the performance-modeling system. The learning algorithms assist PUNCH, a network-computing system at Purdue University, in emulating an ideal user in terms of its resource management and usage policies. 1. Introduction It is now recognized that the heterogeneous nature of the network-computing environment cannot be effectively exploited without some form of adaptive or demand-driven resource management (e.g., [10, 11, 12, 14, 18, 27]). A demand-driven resource management system can be characterized by its a...

: The Purdue University Network Computing Hubs of Indiana, or PUNCH(I), is an Internetbased software infrastructure under development that enables students and faculty at all Purdue campuses in the state of Indiana to use and share unique computational resources. It leverages an existing software infrastructure, the Purdue University Network Computing Hubs (PUNCH). This paper explains the need for a statewide computing system, describes the environment in which PUNCH(I) will be used, outlines the requirements and the goals of an enterprise-wide computing system, and describes how they are met through PUNCH. Key transaction types (in addition to document delivery) supported by PUNCH are described and their performance is discussed. 1. INTRODUCTION The Purdue University Network Computing Hubs of Indiana, or PUNCH(I) is an Internet-based software infrastructure under development that enables students and faculty at all Purdue campuses in the state of Indiana to use and share uniq...