D. Marr. "Artificial Intelligence - A Personal View". Artificial Intelligence 9 (1977), 37-48.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....that some schemes, intended for spatio temporal representation and inference, offer over simple linear inequalities and their algorithms. As such, this work belongs in the implementation slot of Marr s classification of AI research into the levels of computation, algorithm, and implementation [1]. Here, any discussion at the other levels in Marr s triad serves to point out that some current proposals at the implementational level were motivated by a particular task or computation to be achieved. 2. Time and Space in the Abstract Many real world applications of artificial intelligence ....

D. Marr. "Artificial Intelligence - A Personal View". Artificial Intelligence 9 (1977), 37-48.

....that some schemes, intended for spatio temporal representation and inference, offer over simple linear inequalities and their algorithms. As such, this work belongs in the implementation slot of Marr s classification of AI research into the levels of computation, algorithm, and implementation [1]. Here, any discussion at the other levels in Marr s triad serves to point out that some current proposals at the implementational level were motivated by a particular task or computation to be achieved. 2. Time and Space in the Abstract Many real world applications of artificial intelligence ....

D. Marr. "Artificial Intelligence - A Personal View". Artificial Intelligence 9 (1977), 37-48.

....of each concept. Ironically, this violates a venerable adage of knowledge base design: Since one does not usually know in advance what aspect of an object or action is important, it follows that most of the time, a given object will give rise to several different coarse internal descriptions [Mar77] Views facilitate maintaining multiple descriptions of a concept without losing coherence. Each distinct view provides a single, coherent description; the view selection mechanism facilitates activating the most appropriate description for any particular situation. 2. Frames include only ....

D. Marr. Artificial intelligence -- a personal view. Artificial Intelligence, 9:37--48, 1977.

....processing system, when visualised as a network, there are many more fibres coming back from the visual cortex into the Lateral Geniculate Nucleus (LGN) than there are fibres going from the retina to the LGN in the correct direction. How does the computationalist make sense of this Marr (in (Marr 1977), reprinted in (Boden 1990) classifies AI theories into Type 1 and Type 2, where a Type 2 theory can only solve a problem by the simultaneous action of a considerable number of processes, whose interaction is its own simplest description. It would seem that type 2 systems can only be decomposed ....

D.C. Marr. Artificial intelligence: A personal view. Artificial Intelligence, 9:37--48, 1977.

....processing system, when visualised as a network, there are many more fibres coming back from the visual cortex into the Lateral Geniculate Nucleus (LGN) than there are fibres going from the retina to the LGN in the correct direction. How does the computationalist make sense of this Marr (in (Marr, 1977), reprinted in (Boden, 1990) classifies AI theories into Type 1 and Type 2, where a Type 2 theory can only solve a problem by the simultaneous action of a considerable number of processes, whose interaction is its own simplest description. It would seem that type 2 systems can only be decomposed ....

Marr, D. C. (1977). Artificial Intelligence: a personal view. Artificial Intelligence, 9:37-48.

....an image and a model. Groups of features contain a good deal more information than individual features; we can use this information to determine correct matches efficiently. It has long been recognized, however, that grouping is a difficult problem, which perhaps explains its relative neglect. Marr[37] said: The figure ground problem may not be a single problem, being instead a mixture of several subproblems which combine to achieve figural separation, just as the different molecular interactions combine to cause a protein to fold. There is in fact no reason why a solution to the ....

Marr, D., 1977, "Artificial Intelligence - A Personal View," Artificial Intelligence, 9:37-48.

....Mos81, Mar76a, New65, Len82] have only provided syntactic methods without the accompanying semantics. This thesis gives meaning to shifts in formulation by examining the shift in conceptualization it entails. By equating reformulation with reconceptualization we provide a clean Type 1 [Mar76b] theory of reformulation; i.e. we separate an account of what reformulation is, from how to do it. Before we describe what reconceptualizations are, we begin with some intuitions about, and a formalization of, conceptualizations. 2.2.1 Conceptualizations A conceptualization[GN87] is a model of ....

D. Marr. Artificial intelligence: a personal view. Technical Report AIM 355, AI laboratory, MIT, 1976. BIBLIOGRAPHY 130

....biological approach is applied, many of the assumptions of traditional AI have to be reviewed. 1 Brevity requires an element of caricature, but these assumptions are clearly expounded in the work of Newell and Simon [34] figured strongly in the work of David Marr and his followers (e.g. [27, 28]) and are clearly exhibited in most undergraduate AI textbooks I am familiar with (e.g. 35, 43, 8] 3 Biological Issues: Adaptive Behavior Appeals to biology are not new: in the 1960 s [36] and again since the mid 1980 s [37, 29, 1] there has been widespread interest within AI in so called ....

D. Marr. Artificial intelligence: A personal view. Artificial Intelligence, 9:37--48, 1977.

....which reasons on the basis of this world model; passing on its decisions to an action module which translates them into the necessary motor actions. This functional decomposition has been challenged, and an alternative behavioural decomposition proposed, by Brooks in, e.g. Brooks 1991) Marr (in (Marr 1977), reprinted in (Boden 1990) classifies AI theories into Type 1 and Type 2, where a Type 2 theory can only solve a problem by the simultaneous action of a considerable number of processes, whose interaction is its own simplest description. It would seem that Type 2 systems can only be decomposed ....

D.C. Marr. Artificial intelligence: A personal view. Artificial Intelligence, 9:37--48, 1977.

....of what Hendriks Jansen calls interactive emergence [30] between the mechanism and its environment, and is not caused by any discrete functional entity in the former. These two ways of producing behavioural regularities correspond to Marr s distinction between Type I and Type II mechanisms [37]. A rule following, Type I, system has at least two levels of organisation: the lower level of the underlying mechanism, and the higher level of the rules that it follows. For example, a digital computer is realised in an electrodynamical system that displays higher level organisation in the form ....

D. Marr. Artificial intelligence: A personal view. In M. Boden, editor, The Philosophy of Artificial Intelligence. Oxford University Press, 1977.

....One of the strongest early proponents of this view was David Marr, who noted that the perception of an event or object must include the simultaneous computation of several different descriptions of it, that capture diverse aspects of the use, purpose or circumstances of the event or object. (Marr 1977, p. 44) Recently, significant steps have been taken toward representational flexibility with the advent of sophisticated connectionist models whose distributed representations are highly contextdependent (Rumelhart McClelland 1986) In these models, there are no representational primitives in ....

Marr, D. (1977). Artificial intelligence---a personal view. Artificial Intelligence, 9: 37-48.

....[17] Artificial intelligence is the study of complex information processing problems that often have their roots in some aspect of biological information processing. The goal of the subject is to identify interesting and solvable information processing problems, and solve them. Marr, [18]) AI is concerned with methods of achieving goals in situations in which the information available has a certain complex character. The methods that have to be used are related to the problem presented by the situation and are similar whether the problem solver is human, a Martian, or a computer ....

D. Marr. Artificial intelligence: a personal view. Artificial Intelligence, 9:37--48, 1977.

....to the constraints and the components, and selection and activation of the associations. But this language is much too specific to R1 s method. This method specificness of the language becomes a major problem for describing and explaining multi strategy process models such as Kritik3. Task level [Marr 1977] (or, equivalently, knowledge level [Newell 1982] accounts make a clearer separation between knowledge based reasoning and its implementation in a knowledge system. In the mid eighties, Chandrasekaran [1988] proposed the language of Generic Tasks for analyzing and modeling knowledge based problem ....

D. Marr. Artificial Intelligence --- A Personal View. Artificial Intelligence, 9(1), 1977.

....of answer to look for and can be used, often rather decisively, to disprove false theories. 17, p. 132] Crick s doubts over the existence of deep principles of brain function can be phrased in terms of Marr s division of studies of biological information processing into Type 1 and Type 2 theories [40]. Briefly, Marr defines a Type 1 theory as one which accounts for a problem by decomposing it into a set of computations which must be performed: Type 1 theories are little concerned with algorithmic details of implementing the computation; they are prealgorithmic in that they focus on what any ....

....it s conceivable that it could take just as long to get to the bottom of it. McCarthy, quoted by Johnson [32, p.13] The implication is clear: we should cease to concentrate exclusively on high level cognitive function and get the basics right first. Such an argument was also proposed by Marr [40], an advocate of low level studies of human cognition. Marr defended the focus on low level intellectual function by noting that the lower levels are: often dismissed with contempt by those who purport to study higher, more central problems of intelligence. Our reply to such criticism is ....

[Article contains additional citation context not shown here]

D. Marr. Artificial intelligence: A personal view. Artificial Intelligence, 9:37--48, 1977.

No context found.

D.Marr. Artificial Intelligence --- A Personal View. Artificial Intelligence, 9(1), 1977.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC