摘要:

An important issue to consider when applying machine learning to real world problems is the selection of an appropriate learning tool from the large set of available techniques. Building on our experience with the Machine Learning Toolbox, we propose a set of taxonomies that allow a domain expert, with little or no knowledge of machine learning, to choose a suitable tool for his particular application. Unlike previous classifications of learning systems, which were based on technical characteristics of these systems, ours relies on features of the applications that can be solved, such as the user's goal, available data and background knowledge, and interaction between the system and its user.

ISSN:0883-9514    

DOI:10.1080/08839519408945431

出版商:Taylor & Francis Group    

年代:1994

数据来源: Taylor

摘要:

In this paper, a methodology for document classification and understanding is proposed. It is based on a multistrategy approach to learning from examples. By document classification, we mean the process of identification of the particular class to which a document belongs. Document understanding is defined as the process of detecting the logical structure of a document. The multistrategy approach for document classification and understanding has been implemented in a system called PLRS, which embeds two empirical learning systems: RES and INDUBIIH. Given a set of documents whose layout structure has already been detected and such that the membership class has been defined by the user, RES generates the knowledge base of an expert system devoted to the classification of a document. The language used to describe both the layout of the training documents and the learned rules is a first-order language. The learning methodology adopted for the problem of learning classification rules integrates both a parametric and a conceptual learning method. As to the problem of document understanding, INDUBIIH can be used to generate the recognition rules, provided that the user is able to supply examples of the logical structure. RES and INDUBIIH are implemented in C language. PLRS is a module oflBIsys, a software environment for office automation distributed by Olivetti.

ISSN:0883-9514    

DOI:10.1080/08839519408945432

出版商:Taylor & Francis Group    

年代:1994

数据来源: Taylor

摘要:

The system presented, LEFT, learns most specific generalizations (MSGs)from structural descriptions. The new inductive multistaged generalization algorithm is based on several new or enhanced ideas that improve the quality of generalization using weighted predicates and make it applicable to real world problems. The algorithm distinguishes between important and less important predicates. Built-in predicates are used to select alternative MSGs and improve the resulting hypothesis. The system has been applied successfully to chip-floorplanning, a subtask of VLSI design. It acquires rules describing single floorplanning steps.

ISSN:0883-9514    

DOI:10.1080/08839519408945433

出版商:Taylor & Francis Group    

年代:1994

数据来源: Taylor

摘要:

This paper presents the results of a feasibility study on the application of machine learning to knowledge acquisition in transportation engineering. An eight-stage knowledge acquisition process is proposed and its individual stages justified and described. Machine learning is used to learn about urban rail control. The development of the representation space for this problem is discussed, including the analysis of motion and stopping regime for a train, and of both decision and performance attributes. Travel time, energy consumption, and passenger comfort are used as performance attributes. Six automated knowledge acquisition processes were conducted for various performance (dependent) attributes, taking into consideration two different clusterings of performance attribute values into three and seven subranges. All the examples used for learning were computer generated, using REGIME, which separately produces estimations of individual performance attributes for a given train-driving scenario and an assumed rail corridor. The decision rules produced are discussed, and their verification, based on the overall empirical error rate, is reported. This paper also contains conclusions and suggestions regarding future research on applications of machine learning to knowledge acquisition in transportation engineering.

ISSN:0883-9514    

DOI:10.1080/08839519408945434

出版商:Taylor & Francis Group    

年代:1994

数据来源: Taylor

摘要:

We applied inductive classification techniques to data collected in a Boeing plant with the I goal of uncovering possible flaws in the manufacturing process. This application led us to explore two aspects of classical decision tree induction: (1) preprocessing and postprocessing,and (2) brute-force induction. For preprocessing and postprocessing, much of our effort was focused on the preprocessing of raw data to make it suitable for induction and the postprocessing of learned rules to make them useful to factory personnel. For brute-force induction, in contrast with standard methods, which perform a greedy search of the space of decision trees', we formulated an algorithm that conducts an exhaustive, depth-bounded search for accurate predictive rules. We demonstrate the efficacy of our approach with specific examples of learned rules and by quantitative comparisons with decision tree algorithms (C4 and CART).

ISSN:0883-9514    

DOI:10.1080/08839519408945435

出版商:Taylor & Francis Group    

年代:1994

数据来源: Taylor

摘要:

This paper presents a new approach to face recognition, combining the techniques of computer vision and machine learning. A steady improvement in recognition performance is demonstrated. It is achieved by learning individual faces in terms of the local shapes of image boundaries. High-level facial features, such as nose, are not explicitly used in this scheme. Several machine learning methods are tested and compared. The overall objectives are formulated as follows: Classify the different tasks of “face recognition” and suggest an orderly terminology to distinguish between them. Design a set of easily and reliably obtainable descriptors and their automatic extraction from the images. Compare plausible machine learning methods; tailor them to this domain. Design experiments that would best reflect the needs of real world applications, and suggest a general methodology for further research. Perform the experiments and compare the performance.

ISSN:0883-9514    

DOI:10.1080/08839519408945436

出版商:Taylor & Francis Group    

年代:1994

数据来源: Taylor

ISSN:0883-9514    

DOI:10.1080/08839519408945430

出版商:Taylor & Francis Group    

年代:1994

数据来源: Taylor