Rule-Based Expert Systems and Linear Models: An Empirical Comparison of Learning-By-Examples Methods*

Building models of expert decision-making behavior from examples of experts’ decisions continues to receive considerable research attention. In the 1960's and 70's, linear models derived by statistical methods were studied extensively. More recently, rule-based expert systems derived by induction algorithms have been the focus of attention. Few studies compare the two approaches. This paper reports on a study that compared linear models derived by logistic regression with rule-based systems produced by two induction algorithms—ID3 and the genetic algorithm. The techniques performed comparably in modeling the experts at one task, graduate admissions, but differed significantly at a second task, bidder selection.     

Notes and Communications REPLANNING FREQUENCIES FOR MASTER PRODUCTION SCHEDULES

Updating production plans typically is achieved by rolling the planning horizon forward one period at a time, each time including the latest information in order to determine the best course of action to pursue in the present period. Theoretical planning-horizon studies have identified the conditions by which the production decisions in the current and some specified number of future periods remain optimal given some set of future demands. Motivated by these findings, this study addresses the replanning frequency in a hierarchical production planning problem where no planning-horizon theorems are available. In this problem the aggregate production plan and the master production schedule are linked by a rolling-horizon practice. Empirical experimentation indicates that under certain cost and demand conditions the master production schedule need not be updated every period. If a schedule does not need to be updated for several periods, the schedule for these periods can be frozen to provide stability for planning components at lower levels in the bill of material of the products. The results of this study thus provide some reference for the determination of the frozen portion of the master production schedule.     

Child welfare policy and services in Korea

Over the past 20 years, the Korean government has developed child welfare laws to respond to an increasing number of child maltreatment cases and create a centralized child welfare delivery system. However, the system remains fragmented and under-developed due to a lack of understanding of child maltreatment and limited government support. Further, few studies have comprehensively reviewed the system. This study examines the Korean child welfare system, identifies challenges for child welfare policies and services, and discusses the implications for future policy; the results can be used to improve the safety, permanency, and well-being of children and families in Korea. The analysis shows that the government has prioritized family preservation over removal from families, even in substantiated cases of abuse. However, ensuring the safety of children in Korea will require not only working to preserve families but also developing and implementing better options for in-home services and out-of-home placement.     

Robustly assigning unstable items

Journal of Combinatorial Optimization - We study the robust assignment problem where the goal is to assign items of various types to containers without exceeding container capacity. We seek an...