PBD SYSTEMS: WHEN WILL THEY EVER LEARN?
Published 1 January 1996
Ian H. Witten
Citations5
Generate an AI Snapshot to get a quick, structured summary of this paper.
Study Snapshot
ObjectiveStudy objective
MethodsResearch methodology
PopulationPopulation studied
Sample sizeSample sizes
OutcomesStudy outcomes here
ResultsStudy results comes here
LimitationsResearch study limitations comes here
A concise AI-generated summary of the paper will appear here once you click Generate AI Snapshot.
TL;DR
The tragicomic courtship between programming by demonstration and machine learning, two fields that seem made for each other but have never quite got together, is traced and the long-term prognosis is not so much wedded bliss as assimilation.
Abstract
. We trace the tragicomic courtship between programming by demonstration and machine learning, two fields that seem made for each other but have never quite got together. A long-term historical perspective illuminates the twin roles of interaction and sequence learning in programming by demonstration, and reveals the parallel growth of machine learning as a research area in its own right. A view of the present shows a few---but only a few---intimations that these two fields are beginning to develop a meaningful relationship. The long-term prognosis is not so much wedded bliss as assimilation---ultimately PBD and ML shall, in the words of Genesis, cleave unto each other: and they shall be one flesh. Key Words. Programming by demonstration, Machine learning, Sequence learning, Instructability, MDL principle The field of programming by demonstration (PBD) was born in a workshop held at Apple Computer in March 1992, and is effectively defined by the eighteen computer implementations docum...
Keywords
Computer Science
C4.5: Programs for Machine Learning
23,665 Citations1992J. R. Quinlan
A complete guide to the C4.5 system as implemented in C for the UNIX environment, which starts from simple core learning methods and shows how they can be elaborated and extended to deal with typical problems such as missing data and over hitting.
Programs for Machine Learning
5,794 Citations1994Steven L. Salzberg, Alberto M. Segre
In his new book, C4.5: Programs for Machine Learning, Quinlan has put together a definitive, much needed description of his complete system, including the latest developments, which will be a welcome addition to the library of many researchers and students.
A theory of the learnable
4,242 Citations1984Leslie G. Valiant
This paper regards learning as the phenomenon of knowledge acquisition in the absence of explicit programming, and gives a precise methodology for studying this phenomenon from a computational viewpoint.
The Intentional Stance
3,862 Citations1987Daniel C. Dennett
Information and ControlLanguage identification in the limit
3,569 Citations1967Eric Gold
It was found that theclass of context-sensitive languages is learnable from an informant, but that not even the class of regular languages is learningable from a text.
IEEE Transactions on CommunicationsData Compression Using Adaptive Coding and Partial String Matching
1,250 Citations1984John J. Cleary, Ian H. Witten
This paper describes how the conflict can be resolved with partial string matching, and reports experimental results which show that mixed-case English text can be coded in as little as 2.2 bits/ character with no prior knowledge of the source.
Machine LearningLearning Efficient Classification Procedures and Their Application to Chess End Games
1,106 Citations1983J. R. Quinlan
A series of experiments dealing with the discovery of efficient classification procedures from large numbers of examples is described, with a case study from the chess end game king-rook versus king-knight.
Watch what I do: programming by demonstration
1,024 Citations1993Allen Cypher, Daniel C. Halbert +5 more
Part 1 Systems: Pygmalion tinker a predictive calculator rehearsal world smallStar peridot metamouse TELS eager garnet the Turvy experience chimera the geometer's sketchpad tourmaline a history-based macro by example system mondrian triggers the AIDE project.
DSpace@MIT (Massachusetts Institute of Technology)Learning Structural Descriptions From Examples
1,007 Citations1970Patrick Henry Winston
International Journal of Man-Machine StudiesPRISM: An algorithm for inducing modular rules
652 Citations1987Jadzia Cendrowska
A new algorithm is described, PRISM, which, although based on ID3, uses a different induction strategy to induce rules which are modular, thus avoiding many of the problems associated with decision trees.
Munich Personal RePEc Archive (Ludwig Maximilian University of Munich)Version spaces: an approach to concept learning.
341 Citations1979Tom M. Mitchell
The version space approach has been implemented as one component of the Meta-DENDRAL program for learning production rules in the domain of chemical spectroscopy and proofs are given for the correctness of the method for representing version spaces, and of the associated concept learning algorithm, for any countably infinite concept description language.
CERN Document Server (European Organization for Nuclear Research)Creating user interfaces by demonstration
244 Citations1988Brad A. Myers
Peridot, a new User Interface Management System, addresses this problem by allowing the user interface designer to demonstrate how the input devices should be handled by giving an example of the interface in action.
PYGMALION: A Creative Programming Environment
197 Citations1975David Canfield Smith
PYGMALION is a two-dimensional, visual programming system implemented on an interactive computer with graphics display that is an interactive remembering editor for icons, which executes and saves operations for later reexecution.
Programming by example
173 Citations1984Daniel C. Halbert
This research has demonstrated that programming by example is a practical method for creating programs and can be applied to other systems as well.
Unbounded length contexts for PPM
130 Citations2002John G. Cleary, William J. Teahan +1 more
A new algorithm is described, PPM*, which exploits contexts of unbounded length and reliably achieves compression superior to PPMC, although the current implementation uses considerably greater computational resources (both time and space).
Artificial IntelligenceOn the inference of Turing machines from sample computations
84 Citations1972Alan W. Biermann
An algorithm is presented which when given a complete description of a set of Turing machine computations finds a Turing machine which is capable of doing those computations.
International Journal of Man-Machine StudiesBehaviour/structure transformations under uncertainty
70 Citations1976Brian R. Gaines
This paper analyses the problem of determining a structure for an automaton, optimal in some sense, from observations of its behaviour which are themselves uncertain, and shows that appropriate measures lead to the poorness-of-fit of admissible models of a probabilistic source being an entropy for that source.
Machine LearningDiscrete Sequence Prediction and Its Applications
68 Citations1994Philip Laird, Ronald Saul
This work presents a simple and practical algorithm (TDAG) for discrete sequence prediction based on a text-compression method that limits the growth of storage by retaining the most likely prediction contexts and discarding less likely ones.
The Reactive Keyboard
61 Citations2010John J. Darragh, Ian H. Witten
This book describes the Reactive Keyboard, a system that greatly speeds communication by predicting the user's next response before it is made, although it does not always predict correctly.
DSpace@MIT (Massachusetts Institute of Technology)Justified generalization : acquiring procedures from examples
41 Citations1984Peter Andreae
PRISM (University of Calgary)Instructible agents
32 Citations1994David Maulsby
This dissertation explores the design of inference and interaction methods that would enable end users to teach a computer the characteristic, discriminating features of a set of data, and implements the first concept learning system whose dynamic bias enables it to learn from examples, ambiguous hints and partial specifications.
PRISM (University of Calgary)DEMONSTRATION INSTEAD OF PROGRAMMING: FOCUSSING ATTENTION IN ROBOT TASK ACQUISITION
25 Citations1989Rosanna Heise
Elsevier eBooksON ASKING THE RIGHT QUESTIONS
19 Citations1988Brent J. Krawchuk, Ian H. Witten
Different approaches to selecting examples are reviewed, and the ramifications of one in detail are explored, to provide a well-understood theoretical underpinning for concept learning systems.
International Journal of General SystemsAPPROXIMATE, NON-DETERMINISTIC MODELLING OF BEHAVIOUR SEQUENCES
16 Citations1979Ian H. Witten
A technique for forming approximate, non-deterministic, finite-memory models of behaviour sequences that depend on a “memory-length” parameter k which controls their complexity and hence goodness-of-fit lo the given behaviour by constraining them to reproduce correctly all sub-sequences of length k and less in the behaviour.
PRISM (University of Calgary)PROGRAMMING BY EXAMPLE FOR THE CASUAL USER: A CASE STUDY
7 Citations1981Ian H. Witten
University of Canterbury Research Repository (University of Canterbury)Designing teachable robots
7 Citations1984Bruce A. MacDonald
University of Canterbury Research Repository (University of Canterbury)An associative and impressible computer.
6 Citations1980John G. Cleary
Research Commons (University of Waikato)Interactive concept learning for end-user applications
1 Citations1995David Maulsby, Ian H. Witten
The Cima learning system combines evidence from examples, task knowledge and user hints to form Disjunctive Normal Form (DNF) rules for classifying, generating or modifying data.