Eugene Fink

Integrating planning and learning: the PRODIGY architecture

Abstract Planning is a complex reasoning task that is well suited for the study of improving performance and knowledge by learning, i.e. by accumulation and interpretation of planning experience. PRODIGY is an architecture that integrates planning with multiple learning mechanisms. Learning occurs at the planners decision points and integration in PRODIGY is achieved via mutually interpretable knowledge structures. This article describes the PRODIGY planner, briefly reports on several learning modules developed earlier along the project, and presents in more detail two recently explored methods to learn to generate plans of better quality. We introduce the techniques, illustrate them with comprehensive examples, and show preliminary empirical results. The article also includes a retrospective discussion of the characteristics of the overall PRODIGY architecture and discusses their evolution within the goal of the project of building a large and robust integrated planning and learning system.

Selection of patients for clinical trials: an interactive web-based system

The purpose of a clinical trial is to evaluate a new treatment procedure. When medical researchers conduct a trial, they recruit participants with appropriate health problems and medical histories. To select participants, they analyze medical records of the available patients, which has traditionally been a manual procedure. We describe an expert system that helps to select patients for clinical trials. If the available data are insufficient for choosing patients, the system suggests additional medical tests and finds an ordering of the tests that reduces their total cost. Experiments show that the system can increase the number of selected patients. We also present an interface that enables a medical researcher to add clinical trials and selection criteria without the help of a programmer. The addition of a new trial takes 10-20 min, and novice users learn the functionality of the interface in about an hour.

Fundamentals of restricted-orientation convexity

Abstract A restricted-orientation convex set , also called an O-convex set, is a set of points whose intersection with lines from some fixed set is empty or connected. The notion of O-convexity generalizes standard convexity and orthogonal convexity. We explore some of the basic properties of O-convex sets in two and higher dimensions. We also study O-connected sets, which are restricted O-convex sets with several special properties. We introduce and investigate restricted-orientation analogs of lines, flats, and hyperplanes, and characterize O-convex and O-connected sets in terms of their intersections with hyperplanes. We then explore properties of O-connected curves; in particular, we show when replacing a segment of an O-connected curve with a new curvilinear segment yields an O-connected curve and when the catenation of several curvilinear segments forms an O-connected segment. We use these results to characterize an O-connected set in terms of O-connected segments, joining pairs of its points, that are wholly contained in the set. We also identify some of the major properties of standard convex sets that hold for O-convexity. In particular, we establish the following results: The intersection of a collection of O-convex sets is an O-convex set; every O-connected curvilinear segment is a segment of some O-connected curve; for every two points of an O-convex set, there is an O-convex segment joining them that is wholly contained in the set.

Indexing of time series by major minima and maxima

We describe a technique for fast compression of time series and indexing of compressed series. We have tested it on three data sets: stock prices, air and sea temperatures, and wind speeds.

Compression of time series by extracting major extrema

We formalise the notion of important extrema of a time series, that is, its major minima and maxima; analyse the basic mathematical properties of important extrema; and apply these results to the problem of time-series compression. First, we define the numeric importance levels of extrema in a series, and present algorithms for identifying major extrema and computing their importances. Then, we give a procedure for fast lossy compression of a time series at a given rate, by extracting its most important minima and maxima, and discarding the other points.

Scheduling with Uncertain Resources: Search for a Near-Optimal Solution

We describe a system for scheduling a conference based on incomplete information about available resources and scheduling constraints. We explain the representation of uncertain knowledge, describe a local-search algorithm for generating near-optimal schedules, and give empirical results of automated scheduling under uncertainty.

Scheduling with Uncertain Resources: Representation and Utility Function

We describe the representation of uncertain knowledge in a conference-scheduling system, which may include incomplete information about available resources, conference events, and scheduling constraints. We then explain the use of this incomplete knowledge in the evaluation of schedule quality.

Restricted-orientation convexity

1 Introduction.- 1.1 Standard Convexity.- 1.2 Ortho-Convexity.- 1.3 Strong Ortho-Convexity.- 1.4 Convexity Spaces.- 1.5 Book Outline.- 2 Two Dimensions.- 2.1 O-Convex Sets.- 2.2 O-Halfplanes.- 2.3 Strongly O-Convex Sets.- 3 Computational Problems.- 3.1 Visibility and Convexity Testing.- 3.2 Strong O-Hull.- 3.3 Strong O-Kernel.- 3.4 Visibility from a Point.- 4 Higher Dimensions.- 4.1 Orientation Sets.- 4.2 O-Convexity and O-Connectedness.- 4.3 O-Connected Curves.- 4.4 Visibility.- 5 Generalized Halfspaces.- 5.1 O-Halfspaces.- 5.2 Directed O-Halfspaces.- 5.3 Boundary Convexity.- 5.4 Complementation.- 6 Strong Convexity.- 6.1 Strongly O-Convex Sets.- 6.2 Strongly O-Convex Flats.- 6.3 Strongly O-Convex Halfspaces.- 7 Closing Remarks.- 7.1 Main Results.- 7.2 Conjectures.- 7.3 Future Work.- References.

SmartNotes: Application of crowdsourcing to the detection of web threats

We describe a crowdsourcing system, called SmartNotes, which detects security threats related to web browsing, such as Internet scams, deceptive sales of substandard products, and websites with intentionally misleading information. It combines automatically collected data about websites with user votes and comments, and uses them to identify potential threats. We have implemented it as a browser extension, which is available for free public use.

Scheduling with Uncertain Resources: Elicitation of Additional Data

We consider the task of scheduling a conference based on incomplete data about available resource and scheduling constraints, and describe a procedure for automated elicitation of additional data. This procedure is part of an interactive system for scheduling under uncertainty, which identifies critical missing information, generates related questions to the human administrator, and uses answers to improve the schedule.