Date of Submission

10-28-1994

Date of Award

10-28-1995

Institute Name (Publisher)

Indian Statistical Institute

Document Type

Doctoral Thesis

Degree Name

Doctor of Philosophy

Subject Name

Computer Science

Department

Machine Intelligence Unit (MIU-Kolkata)

Supervisor

Pal, Sankar Kumar (MIU-Kolkata; ISI)

Abstract (Summary of the Work)

Machine recognition [1, 2] of patterns can be viewed as a two-fold task, consisting of learning the invariant and common properties of a set of samples characterizing a class, and of deciding a new sample as a possible member of the class by noting that it has properties common to those of the set of samples. In other words, pattern recognition by computers can be described as a transformation from the measurenment space M to the feature space F and finally to the decision space D (1), i.e., M ⟶F⟶D.Here, the mapping 6 : F⟶D is the decision function and the elements d D are termed as decisions.When the input pattern is an image, the measurement space involves processing tasks such as enhancement, filtering, contour extraction and noise reduction, with the objective being to extract salient features from the pattern. This is termed as image processing [3, 4). The ultimate goal is the understanding, recognition and interpre- tation using the processed information available from the image pattern. A complete image recognition/ interpretation scheme is called a vision system (5, 6] and can be viewed as consisting of three levels, viz., low level, mid level and high level.Artificial intelligence is the field that investigates how computers can be made to exhibit intelligence in different aspects of thinking, reasoning, perception or action. In other words, it involves the study of the mental faculties using computational models (7). A key observation in this direction is that knowledge should not be represented in heavily encoded forms that suppress the structure and constraints. This has lead to the development of more explicit, symbolic, highly flexible forms of representation of knowledge, such as semantic nets, frames and production rules which can be efficiently processed. An expert system can be viewed as a rule-based application program which provides the user with the facility for posing and obtaining answers (that require expertise) to questions related to the information stored in its knowledge base (8, 9]. Basically it consists of the knowledge base, the inference engine and an user interface linking the external environment to the system. The model typically functions in a narrow domain dealing with specialized knowledge generally possessed by human experts. Such systems possess a non-trivial inferential capability and are expected to be capable of directing the acquisition of new information in an efficient manner. The knowledge base of an expert system is problem-dependent and contains information that controls inferencing. Traditional rule-based models encode this information as If- Then rules. In Classification expert systems (10], the outputs are represented by class variables each of which can assume continuous values. Programs for diagnosis, fault detection and pattern recognition are examples of applications that can be represented as classification problems and are handled by this type of expert systems. In the medical domain it can be used for diagnosing a particular symptom set as affliction by one or more disease(s).Fuzzy sets were introduced in 1965 by Zadeh [11] as a new way of representing vagueness in everyday life.

Comments

ProQuest Collection ID: http://gateway.proquest.com/openurl?url_ver=Z39.88-2004&rft_val_fmt=info:ofi/fmt:kev:mtx:dissertation&res_dat=xri:pqm&rft_dat=xri:pqdiss:28842985

Control Number

ISILib-TH272

Creative Commons License

Creative Commons Attribution 4.0 International License
This work is licensed under a Creative Commons Attribution 4.0 International License.

DOI

http://dspace.isical.ac.in:8080/jspui/handle/10263/2146

Download

Included in

Mathematics Commons

Share

COinS