Saturday, November 12, 2005

Meclazine And Vertigo



One of the practical applications of Artificial Neural Networks (ANN), is the classification of data, understood as a process of searching for properties common to a number of objects a domain of knowledge, depending on the values \u200b\u200bof certain attributes. Within the issue of automatic qualification, while the alternative process of more general processes encompassed within what is known as "machine learning", one of the learning algorithms known, based on "examples", is called ID3, or "Iterative dichotomized (version) 3" (JR Quinlan, 1979). Work with symbolic data, as opposed numerical data, and is based on obtaining a decision tree (see Annex
  • ), from which derive a set of production rules, can represent a particular domain or universe, producing independent knowledge of that domain (the learning system from an initial state chosen domain in which there is no knowledge base, extracting patterns common among the examples used, from which generates a knowledge base application that domain). The decision tree can therefore classify the input data. We can distinguish two types of learning processes: supervised
  • : examples or explanations "were provided to the system by an external subject. This category ratings data based on decision trees based on examples, such as learning algorithm ID3 . Unsupervised
  • : examples or "comments" are created by the system itself. This category grouping processes data or data clustering (or just clustering ).
  • Based on a conceptual level of abstraction higher, the so-called "machine learning" or learning machine, it is possible to distinguish two types of learning: rote learning

    ID3 of Quinlan.

    Most of the heuristics used for the determination of decision trees through learning algorithms, are based on the mathematical theory of information ( C .

    Shannon, W. Weaver, Bell Laboratories, 1948) [1
    • ] [ 2]. Heuristics are criteria, methods or principles for deciding, from among several alternative courses of action, which will be most effective to achieve certain goal. Restrict the number of evaluations, and therefore impact on improving the search time for solutions. Entropíay amount of information are two concepts that come together in the field of heuristics. About Entropíay amount of information, see Uncle Petros
    • : [1 ] [ 2 ] [
    3] [4

    ]. ID3 algorithm

    generates what is called the rules "hard", ie those who only see two possible states (true or false, positive-negative, 0-1, etc.), And which are both a bivalent character, unlike the rules "fuzzy", which can represent an infinite range of values \u200b\u200bbetween two ends of a scale as those obtained using algorithms ID3

    "extended" (ID4, ID5, ID5R , C4.5, C5, etc.).

    ID3 Algorithm Pseudocode

    :

    If all examples belong to E same class C, then arbol1 \u0026lt;- node labeled with C

    But

    If a = f, then C \u0026lt;- class majority of the examples of E arbol1 \u0026lt;- node labeled with C But A \u0026lt;- best attribute of a arbol1 \u0026lt;- node labeling with A For each v belonging to the values \u200b\u200bof A, do VAS \u0026lt;- examples of E with the value v for attribute A If Eav = f, then ; arbol2 \u0026lt;- node labeled with the majority class in E But arbol2 \u0026lt;- ID3 (VAS, a-{A}) arbol1 \u0026lt;- add to the arbol2 arbol1 through a branch labeled with v arbol1 Return Another representation in pseudocode algorithm ID3 :

    -Decision-Tree Learning (

    Examples, Attributes ,

    Default)

    returns a decision tree

    IF there
    Examples
    , Default
    return


    ELSE IF

    if all examples have the same classification
    ,

    return
    classification,



    ELSE IF s = Attribute

    empty

    return Majority ( Examples

    ) ELSE

    best-atr \u0026lt;- choose-attribute (Attributes , Examples
    )
    tree \u0026lt;- new decision tree rooted in best-atr
    FOR EACH value v [i] best-atr DO Examples
    [i] \u0026lt;- Examples { items with better-atr = v [i]} Subar \u0026lt;--Decision-Tree Learning (examples [i], Attributes
    - best-atr, Majority ( Examples ))
    add branch the tree with label v [i] and subtree Subar OD return tree learning processes that make use of the classification of data by discovery of patterns, are widely used in what is known as "Data Mining", Castilian data mining, data mining or knowledge discovery in databases, terminological diversity about which there is discussion.
    Maximiliano del Rio is the author of a version written in Prolog language learning algorithm ID3
    . The files for this implementation (library Ratings ) can be located either in the source code section of
    programacion.com
    (compressed into a "zip"), or personal space that the author has the "Wiki " of SWI-Prolog
    . In " guia.txt " describes the management of this implementation
    ID3 algorithm in Prolog, which uses the ODBC

    interface to query the tables of the database selected, the examples are obtained for the generation of production rules. Is also attached file " clasif.pl " program "Data Mining" that uses the ID3 algorithm
    endowed GUI using the native library XPCE

    .

    "[...]

    program that uses the above library [...] helps generate the rules and displays the textual and graphical rules obtained, also shows a trace of how the algorithm works. " Source

    This GUI is opened releasing the target "? - Main." on the command line SWI-Prolog, once compiled the program. Finally, the library " compila.pl " contains predicates that allow generate an executable for Windows from the results, using SWI-Prolog. For a rather wide on the Prolog implementation of automatic learning processes in general and inductive learning by decision trees in particular (data classification) is highly recommended reading Chapter 18, "Machine Learning", the (now classic) work of Ivan Bratko, "Prolog: Programming for Artificial Intelligence" (2 nd ed. Addison-Wesley, 1994, ISBN: 0-201-41606 - 9). On decision trees specifically point is 18.6, "Induction of decision trees." There is thus itself a repository of machine learning algorithms written in Prolog, Prolog library of machine learning algorithms , albeit somewhat outdated, since the last update seems to date from 1994, maintained by Thomas Hoppe (Fraunhofer-Gesellschaft , Technical University of Berlin ). The programs are written using the syntax and, in most cases, the predefined predicates (built-in predicates ) specified in Prolog by Clocksin and Mellish described, known as standard Edinburgh, based in turn on the DECSYSTEM-10 (D. Warren, F. Pereira and L. Pereira), to thus ensure the greatest possible degree of compatibility between versions of the language. algorithm implementations ID3 are located in the " RTD" (see in any case the file "Readme " for more information). More information:

    ID3 algorithm JR Quinlan
    (document translated by JA Fernandez,

    PDF, compressed in a zip). Basics of Symbolic Learning (JG Boticario).

    Learning classifiers (Berzal F. Galiano, in PDF). ART: An alternative method for constructing decision trees (Berzal F. Galiano, in PDF). ART: An alternative method for constructing decision trees (Berzal F. Galiano, 2002; doctoral thesis in PDF). RTD Torgos ID3-like system based on the gain-ratio measure ( ID3 algorithm written in Edinburgh Prolog syntax.) This code is located in the directory on Machine Learning

    extensions

    ] [Back to text

    ]

    Friday, November 11, 2005

    Tawneee Stone Wikipedia

    ID3 learning algorithm mnemonics and riddles of wit

  • Note:
  • the first part of this extensive annotation n, the most closely related to the wording of the title, and was posted in my blog sister,
  • Seen and Read, the 16/01/2004

    under the same title. At some later point link that is the application of logic programming in general and Prolog in particular, to the riddles of wit and logic games, I added other issues strictly related to that programming language, particularly regarding its origin, evolution and some of its versions developed over time with less success and permanence maso later.

    In the book "Helping memory. Techniques and tricks for remembering" (Plaza & Janes, 2001, ISBN: 84-8450-527-8) its author, Josep M ª Albaigès, includes, as essential complement to the excellent description in the book is addressed on retention processes, memory, and memory retrieval techniques of saving, a curious "Dictionary of mnemonics, which collected, sorted alphabetically in the issue which they refer, all kinds of mnemonic devices (of

    mnemo
      +
    • markering
    • , those who serve to assist the memory), some rather complex and more affordable, referring to many different aspects. In this same dictionary is the following definition:
    • "We called mnemonics to any trick or shortcut that allows recall procedure, so maso least provisionally, a concrete thing and generally brief."
    • In "Helping memory ...", page 95
    • I stay for the occasion with two curiosities referring to the universe of ratings, they understood in broad sense (I quote):

    Decimal Classification in library The themes of the decimal classification usually: 0. General Works - 1. Philosophy - 2. Religion - 3. Social Studies - 4. Language - 5. Pure Science - 6. Technology - 7. Art - 8. Literature - 9. History. To be remembered with the phrase: "General: religious philosopher! Partners: speak purely! Technicians and artists, historical write!".

    is also very curious description of menotécnica syllogisms (and how to remember the figures where they classified scholastics), the number pi , and several principles or fundamental theorems of physics and statistics, among others, but given the relatively large area of \u200b\u200bthe texts in question, I refer to the query and reading the book referenced, if interested in this topic.

    syllogistic logic or Aristotelian

    trying to determine the truth or falsity of certain philosophical argument, by contrasting propositions or premises , and in a sense can be considered as a formalization No, based on natural language expressions of common sense. The use of the syllogism by Scholastic

    is understandable given the integration of philosophy of Aristotle in Christian dogma characteristic of the task of reflection developed by this group of thinkers in Western philosophy between mid-eleventh century and the mid-fifteenth century or so. Josep M ª

    Albaigès is also editor of Carrollia

    ,

    [...] "[...] press organ of Mensa Spain

    that is devoted to mathematics recreational linguistics, Experimental Literature, Logic, Science and everything would have liked to

    Lewis Carroll. "

    Source


    The

    of recent newsletters years are available in PDF format

    . It is also highly recommended, if you are interested in these issues, the "Stock puzzles" (solved) of Mensa Club

    , most of them originally published in the pages of Carrollia. On another "utility" practice, "

    Official Gazette of the Faculty of Useless (BOFCI) is another publication of Mensa Club whose reading should not be missed ... And while we're at least tangentially, game logic and wit, to mention two blogs in English in a monograph devoted to these issues: Ingenio Games & Puzzles and Small Enigmas you

    . In this latter site is located a good number of links that lead to other pages related themes (see the "Good pages of wit"). Linking

    initially question the use of logic programming in general and Prolog in particular, in solving puzzles and logic games, we can mention mode such excellent work "Solving riddles

    Prolog (
    J. Peri

    , National University of Luján , Argentina, in PostScript format ), where we read:

    In" Solving puzzles with Prolog

    " The examples are written in Micro-Prolog notation , which differs substantially from the subsequently released as a standard facto by the so-called "Edinburgh Prolog, as explained in an appendix at the end of the document (Appendix II - Differences between Micro-Prolog and the notation of Edinburgh), although the semantics execution is identical in both cases. The rationale underlying the examples given is fully valid in any case, regardless of the notation with which they represent. The author of this paper is coordinator of the Group Functional and Logic Programming (UNL

    , UNP

    , Argentina) whose members have also published other interesting documents and papers on various aspects of logic programming, searchable on the website of that group (be patient with the downloads, the server is quite slow, and are delayed for some time). The use of Micro-Prolog notation in the work of this group is explained on the grounds that its members have developed an interpreter for Prolog, Edulog

    oriented towards educational purposes and OPERATION n in their courses and seminars, which just makes use of the notation we have been discussing. The project seems to be paralyzed, at least in the form of public communication, because the pages take a long time without being updated. The prototype

    Edulog is accessible from a link located on page the course "Functional Programming and Logic ", although the download file is password protected. I suppose contacting the project managers, and exposing the reasons why you want to access your application, there will be no problem getting the corresponding key. notation implemented by Micro-Prolog is more limited and less flexible than the standard of Edinburgh itself: for example, any variable names are very restricted, and the interpreter will comply with the first solution that meets the stated objective, as opposed to the usual modern performers based on standard Prolog, which is trying to provide all possible solutions to exhaust the knowledge base program and the possibility of unifying the variables involved. Other differences relate to the use, by Micro-Prolog, predicate particular predefined character. The basic differences between the two types of notations are well explained, with examples, the paper "Differences between notation Micro Edinburgh Prolog and

    " found in the page mentioned subject, although the archive that serves as a means to download the original Word format is also password protected, so again it becomes necessary to contact with their authors to obtain an authorization.

    Micro-Prolog was a dialect of Prolog built around 1980 to be executed originally in the first personal computers, equipment, low processing power and memory compared with the current parameters, based on the Z-processor 80 of Zilog

    Corp., successor to the Intel 8085, and SO
    CP / M

    predecessor of PC / MS-DOS

    (

    ZX Spectrum, Commodore 64 , etc. ). The compiler and interpreter LPA Prolog Professional notation supports Micro-Prolog, although it should be noted that this is a "dialect" completely obsolete, given its obvious limitations, being the standard notation of Edinburgh (also known as syntax-DECSYSTEM 10) the most widespread, and therefore used in almost all current implementations of Prolog language interpreters. Since this is another link can download the original interpreter for Micro-Prolog. WF Clocksin and CS Mellish, in his book "Programming in Prolog" (Gustavo Gili, 1993; ISBN: 84-252-1339-8), devoted an entire section, "Appendix E - Micro-Prolog, to the explanation of the syntax implemented by Micro-Prolog:

    " Micro-syntax Prolog is quite different [...] The basic idea is that there is only one type of term: the list. If we build the 'term' with

    functor f

    and four arguments, we in fact a list of five elements, f as the head, and appearing on four arguments as the remaining elements. Thus, what the syntax of our 'core' is type:

    f (a, g (2.3), c)

    be written in Micro-Prolog: (f (g 2 3) c) . Here we also see different syntax for lists, where lists are enclosed in parentheses, with its elements separated by spaces. clauses are represented as lists of terms, in which the first is the head of the clause, and the rest are goals, which taken as a conjunction, form the body. Here is a more complicated clause [...]: ((alter (z1 (X E - Micro-Prolog Moreover, in Turbo Prolog programs are divided into different sections: domains, clauses, predicates , database and objectives. The explanation, both on the syntax of Micro-Prolog, as on other general aspects of this dialect (predefined predicates particular, facilities for debugging code, etc.), Has some length, so I refer to the referenced site for more information, and in particular, and referred to the Appendix E, Appendices "C - Different versions of Prolog" (provides a brief overview of this particular), and "D - Prolog of DECSYSTEM-10."

    Peter Van Roy (

    Catholic University of Leuven, Belgium,

    Computer Engineering Department ), one of the leading figures in the world of programació n logic, is the author of an article, highly recommended reading, referred to the development of Prolog en la década que transcurre
    a partir de la aparición de la arquitectura WAM :
    " 1983-1993:
    The Wonder Years of Sequential Prolog Implementation
    " (en formato PS ):




    "This report surveys the major developments in sequential Prolog
    implementation during the period 1983-1993. In this decade, implementation
    technology has matured to such a degree that Prolog has left the university
    and become useful in industry. The survey is divided into four parts. The
    first part gives an overview of the important technical developments starting
    with the Warren Abstract Machine (WAM). The second part presents the history
    and the contributions of the major software and hardware systems. The third
    part charts the evolution of Prolog performance since Warren?s DEC-10 compiler.
    The fourth part extrapolates current trends regarding the evolution of
    sequential logic languages, their implementation, and their role in the marketplace."




    "By 1983 Warren had developed the WAM, a structure-copying execution
    model for Prolog that has become the de facto standard implementation technique
    [...]. The WAM defines a high-level instruction set that maps closely to
    Prolog source code. [...]"


    En "

    1983-1993:
    The Wonder Years of Sequential Prolog Implementation

    "


    La tesis doctoral de Peter Van Roy versaba precisamente sobre la implementación
    de the WAM

    : "Can Logic Programming Execute

    as Fast as Imperative Pro-gramming?

    " (PhD Thesis, Department of Computer Science, UC Berkeley, Report UCB / CSD 90/600, 1990).
    The appearance of the Warren Abstract Machine or WAM
    (David HD Warren, 1983) marked a turning point in the evolution of language Prolog, because it entailed the definition of a set of high level instruction and an execution model for this language, the materialization of

    Constraint Logic Programming (Constraint Logic Programming, CLP). Is the basis adopted as a de facto standard for the implementation of portable interpreters for Prolog, making it a multi-language with features similar to compiled languages \u200b\u200bface procedural character, as alos referred runtimes. Thus, in the website of the SWI-Prolog interpreter read: "SWI-Prolog is a Prolog Implementation based on a subset of the WAM (Warren Abstract Machine)." "[...] is based on a very restricted form of the WAM [...] described in Bowen
    & Byrd, 1983

    which defines only 7 instructions. Prolog can
    easily be compiled into this language and the abstract machine code is
    easily decompiled back into Prolog. As it is also possible to wire a standard
    4-port debugger in the

    WAM interpreter there is no need for a distinction between compiled and interpreted
    code. Besides simplifying the design of the Prolog system itself this approach
    has advantages for program development: the compiler is simple and fast,
    the user does not have to decide in advance whether debugging is required
    and the system only runs slightly slower when in debug mode. The price
    we have to pay is some performance degradation (taking out the debugger
    from the
    WAM
    interpreter improves performance by about 20%) and somewhat additional
    memory usage to help the decompiler and debugger."


    "SWI-Prolog extends the minimal set of instructions described in Bowen
    & Byrd, 1983
    to improve performance. While extending this set
    care has been taken to maintain the advantages of decompilation and tracing
    of compiled code. The extensions include specialised instructions for unification,
    predicate invocation, some frequently used built-in predicates, arithmetic,
    and control (
    ;/2 ,
    ( \\ + / 1 ). "

    In

    SWI-Prolog Reference Manual

    - SWI-Prolog 1.1

    The function of this virtual machine, within the core shell , is to compile the instructions written in high level language to machine language instructions, in a similar way as does the Java virtual machine (JVM
    ). WAM architecture has enabled thereby to develop versions of Prolog " compiled "at the implementation level, even if the programs continue to save and run from files that contain code written high-level language. In general, a virtual machine is therefore an abstract procedure to execute sets of instructions written in formal language, such as in this case, Prolog.

    "[...] if we use an interpreted programming environment of Prolog [...] efficiency in terms of execution speed is concerned, decreases, as in any language, to address to a compiled language. The instruction set of modern computers is very low in relation to the semantics of Prolog, so most compilers [...] Prolog perform a clean build an intermediate language rather than directly to machine language. The most popular language is the Warren Abstract Machine (WAM) [...] abstract set of instructions for use in Prolog and can be interpreted or translated into machine language. Other Prolog compilers that translate a high level language such as Lisp or C, and use the compiler to translate into machine language.

    Warren Before working on the compilation of inference in Prolog, logic programming was too slow for widespread use. The Compilers designed by Warren and others allowed to appropriate speeds Prolog [...] More recently, the application of modern technology has allowed compilation [...] Prolog reach optimal speeds as it does compete with C in terms of standard aspects [...]. [...] Being able to write a planner or a natural language parser a few dozen lines of Prolog, make it more preferable than C in the realization of prototypes much of the research projects IA of small scale. "

    Source (

    general frame structure of the page

    cited)

    More information about" Warrens's Abstract Machine: A Tutorial Reconstruction "(in PDF , Hassan Aït-Kaci, The MIT Press, Cambridge, MA, 1991). The following paper describes the system and the notation DECSYSTEM-10 Prolog, which is based the standard of Edinburgh, who remember is the de facto standard Prolog system today, picked up the aforementioned existing ISO this language (ISO / IEC 13211-1:1995 Part 1: General core):

    "

    DECSYSTEM-10 Prolog User's Manual ." DL Bowen (ed.), L. Byrd, FCN Pereira, LM Pereira, and DHD Warren, Department of Artificial Intelligence, University of Edinburgh, November 10, 1982. While the format of the earlier document is. "Doc", this is actually a plain text file, so it disturbs the disposition thereof when opened with MS Word. Also: HTML version, and Castilian translation of this manual, limited to three chapters. The DECSYSTEM-10 Prolog was created and developed an implementation (1977, 1980) for D. Warren and others (F. Pereira, L. Byrd, L. Pereira) in the Artificial Intelligence Department at the University of Edinburgh, to operate computer systems DECSYSTEM-10. Together with the Prolog described in the book referenced in Clocksin and Mellish, versions of the DEC-10 Prolog and its subsequent developments are the basis or standard Edinburgh syntax.

    DECSYSTEM-10 The initials refer to the computer system that was based on the Prolog implementation of the University of Edinburgh we have been discussing, manufactured by Digital Equipment Corp. The
    DECsystem-10/20 systems were 36-bit architecture
    , whose first versions date back to 1971. When we speak of such systems do not have to think about the PC

    s, later in time and initially oriented towards other end users, but was located equipment for processing power and calculation (and all is said, by difficulties in the management and operation), on the scale of large computers known as mainframes. It is not uncommon to find in manuals and guides Antique Prolog (eg those related to C-Prolog [

    HTML] [PS

    in

    ] and general implementations of the first specification derived DECSYSTEM-10 Prolog), references to TOPS -20, DECsystem-10/20 developments, and other systems

    DEC. On the basis of this first Prolog interpreter / compiler was created in 1983, one of the first performers of such language pure, C-Prolog

    (F. Pereira, D. Bowen, L. Byrd), written in C language, which helped to consolidate the standard Edinburgh, as a reference for future implementations. To end this entry, mentioning the example of practical application of Prolog to solve logic games and puzzles under " Solving Rubik's Cube Using the Bestfast Algorithm and Profile tables - A Prolog program and demonstration of an efficient heuristic search method "(by Winston Miller DL
    ). As the title indicates, this text explains the theoretical foundation and operation of a program designed to search for the best possible methods of solving the problem posed by the famous Rubik's Cube

    . The files that make up the program (see end of document) are originally written to be executed under the arity Prolog interpreter (free use) but, by enduring the notacióny

    of predicates is Edinburgh standard, the program in principle should also be able to run properly under SWI-Prolog

    .

    The program uses a search algorithm A * (

    example) based on the type known as "best-first search", which in turn derived from the search method "breadth-first search" or "breadth-first search (example

    ), which evaluates each node the same level, within the search tree (meaning the representation of directed graph-the nodes that are connected by arrows indicating the direction of movement "of a particular search problem), before continuing the process at the next level of depth. In contrast this method lies Search "depth-first" or "depth first" ( example), which explores each "branch" of the tree until you reach "terminal node", before proceeding to search another way. This is precisely the kind of search for "internal" by default holds the Prolog (depth-first search from left to right in And-Or tree) into the compiled knowledge base, the try to meet the objectives. "The A * [...] is an efficient and Widely Used

    pathing algorithm. The A * Keeps track of the cost of getting from point A to point B overalls and the cost of getting from the Beginning point to the end point. The algorithm Follows the path of least cost to find a solution. The A * uses up a lot of memory and Can Be very slow on rough terrain - THEREFORE it is Not Used for real-time robotics

    pathing

    too often. " Source Two basic types search algorithms mentioned (there are many others), they can eventually add heuristic evaluation, consisting of rules able to assess whether, based on certain types of prior information Search is following the right path, for thereby increasing the probability of finding the right solution. Heuristic methods try to apply two criteria: a) choice of the paths that might fail are more likely to like this "pruning" the search space and thus reduce its size, or b) verification, first, the roads most likely to succeed. The heuristic search searches are therefore based on experience. The addition capacity heuristic search methods of "breadth-first", giving rise to algorithms "best-first search", as the A * used in the program applied to the resolution states Rubik's Cube. These three basic types of search algorithms and the concepts and basic theorems of Graph Theory [1 ] [ 2] [3 ] necessary for comprensióny knowledge of their mechanisms of operation, are fully described in Chapter 11. "Basic Problem-Solving Strategies, "12." Best First: A Heuristic Search Principle ", and 13." Reduction and Problem AND / OR Graphs "of the work of Ivan Bratko," Prolog programming for Artificial Intelligence (Addison-Wesley, 1994, ISBN: 0 -201-41606-9).

    Moreover "Artificial Intelligence Through Search

    " (CJ Thorton, B. du Boulay) is a book that specifically addresses the question of methods and algorithms used search space (defined by the set of all nodes in a directed graph). Unfortunately, in the electronic version of this book loses much of original figures and diagrams in print, which makes the understanding of these issues, which in itself has some complexity.

    More information:

    " Games as teaching tools. Formalization of Logic in Prolog Games ( F. Llorens, M ª Jesús Castel, F. Mora and C. Villagra, in PDF ). Turbo Prolog.

    Early History of Prolog: The birth of Prolog / La naissance de Prolog (in PDF , P. Roussel).

    Course Notes on Prolog:

    Chapter 7 Graphs (graph representation in Prolog). Source

    some algorithms A *. , compatible with Visual Prolog

    +

    IDA * (Iterative Deepening A * search), and b) in LPA Win-Prolog syntax . The A * algorithm in Prolog (in

    Prolog Tutorial

    of JRFisher ).

    A * for the Masses
      (in
    • generation5.org ).

    Wednesday, July 20, 2005

    Where Can I Buy Cake Sparklers

    selection of links

    Following the practice introduced months ago on this site, then outlines a series of articles, notes, and links diverse, collected in recent weeks (the quotes are quotes taken from the referenced sites):
  • articles, introductions, annotations of "blogs"
  • Searching for Logic (in Logic and Language ). This entry relates to the application of the laws of logic to the scope of searches in library catalogs.
  • Some Experiences in the Use of CAD System for Teaching Artificial Intelligence
  • (PDF , M. Lezcano, V. Giraldo Valdes in
  • Mathematical Publications, vol. 6 n º 2. , 1998). CAD = Computer Aided Instruction (CAI English, Computer Aided Instruction
  • , or Computer-assisted Instruction ). Practical Common Lisp (on the book by Peter Seibel ; on Slashdot
  • ). linguistics and science fiction (by Xavier Riesco Riquelme in Nessus file). "The only literary genre that has dealt with the language itself constantly, at least in recent years, it is known, the so-called science fiction. Not only from the standpoint of theme a novel, but also from the point of view of literary criticism and essays. Many books of the genre usually contain linguistic elements as subsidiary to the central idea.
  • [...]".
  • An AI tool for the real world - Knowledge modeling with Protégé (H. Knublauch; in
  • JavaWorld, June 2003): "While artificial intelligence (AI) is Often Regarded as an exotic playground academic, Tools and Techniques STI Have to Contribute to mature real-world software technology as well. This article introduce

    . The Resulting models dog Implement decision-support systems, software Requirements capture, populate databases, generate Java classes and

    UML diagrams, share and reuse domain models, and access the Semantic Web. ".

    Progic 2005 ("Second Workshop on Combining Probability and Logic: Special focus on Objective Bayesianism", Centre for Philosophy of Natural and Social Science, London School of Economics, 6-8 July 2005).

    • Library-Centric Software Design LCSD'05 (Workshop to be held in the framework of the conference Object-Oriented Programming, Systems, Languages \u200b\u200band Applications OOPSLA
    • '05, San Diego, California, 16-20 October 2005): "Libraries Are central to all major scientific, engineering, and business areas, yet the design, Implementation, and use of libraries Are underdeveloped arts. This workshop is one of the first steps in the process of Placing all Aspects of libraries on a sound basis Through Technical and Scientific Research into fundamental Issues of best practices and documentation.
    • [...]" A reference to one of the referenced sites (
    • Mathematics and Computation
    • ), I note the simple system in place to provide text content and mathematical symbols ; policy, in effect making use of a JavaScript script called ASCIIMathML , which I will discuss later.

    Saturday, May 7, 2005

    Woman Doctor Inspects Cock

    Knowledge Representation on the Semantic Web

    Within the vast literature published on the Semantic Web
  • [
  • 1] [2
  • ] [
  • 3] [4
  • ] [5
  • ], I this time recommend reading two articles:
  • The Semantic Web (Berners-Lee T., J. Hendler, O. Lassila
  • , in Scientific American

    ) based in frames (frame-based representation ) as a paradigm for building ontology-necessary in its opinion for the construction of the semantic web - and for the conceptualization of formal language

    1] [2

    ] [3 ] [ 4] [5 ] proposed for this purpose as a basic scheme for WWW Consortium as an expression of frame-based representation of one's own WWW

    . and Methodological Issues and

    distribution of knowledge. The knowledge representation scheme based on frames was proposed initially by MIT professor Marvin Minsky -pioneer IA, the neuro-cognitive sciences , and linguistics computing, including disciplines, in his essay A Framework for Representing Knowledge (1974). In any case, as illustrative of the difficulty of the practice of knowledge representation and information retrieval, metadata [ 1] [2 ] to be accurate-in the environment semantic web, nothing better than reading about the article by Peter Manuel Díaz Ortuño Issues and trends in web metadata architecture ( PDF

    , In Journal of Documentation

    , No. 6, 2003, pp. 35-58).

    Monday, April 18, 2005

    Reset Luggage Combination

    Artificial Life: blog

    Through referrals (Referred
    ) of statistics, I washed up in the log Artificial Life (Artificial Intelligence, Software Libre and something of everything else "). In it, among other interesting notes and comments, we find a section dedicated to "hang" Castilian translations of articles originally published in Generation5 (portal dedicated to Artificial Intelligence and multiple perspectives), magnificent opportunity to access its content by those who are not fluent too (or not) the English language, given the good quality of service they usually have, and Orienteering n eminently practical, but it is true that except for an introductory the , are generally intended to be read by people with previous knowledge about the subject in particular. On Language and Prolog Natural Language Processing (NLP )-specific weaknesses, among others, of the undersigned, "there are a few in Generation5 , highly recommended for those wishing to enter the knowledge of these issues.

    Sunday, April 17, 2005

    Birth Control And Aging

    Glossary First Order Logic

    A highly recommended resource, Glossary of First-Order Logic : "This glossary is limited to September basic theory, basic recursive function theory, two branches of logic (truth-functional propositional logic and first-order predicate logic) and Their metatheory. "

    The author of this glossary of terms,

    Peter Suber, teaches in the Department of Philosophy of Earlham College ( U.S. ) Logical Systems course (on formal logical systems, standard first order logic, and Theory of computing, mainly), from whose website can be accessed, under the heading "Hand-outs, a good number of texts in relation to the topics covered in the course. The "hand outs" are materials (booklets with texts and exercises), generally schematic, but not necessarily, delivered to participants of a course. Besides the aforementioned course, P. Suber also developed considerable activity around the open access movement ( Open Access Movement, see your personal page

    , and to highlight something about it, for example your blog Guide to the Open Access Movement

    , among others). It also maintains an extensive directory of links to web Philosophy in : Guide to Philosophy on the Internet . And since we are engaged in logic and philosophy, to recommend Read the following test: Philosophical Logic and Computational Constraints (2002), John L. Taylor, author likewise an interesting blog, Johnny Logic . There is extensive and the reading is quite affordable, providing an overview of the item contained in the title. Through his blog I was able to find some other way on issues of Logic and Philosophy of Logic, for example LogBlog maintained by Richard Zach ( Professor of Philosophy at the University of Calgary, Canada). In turn, the links section (

    Blogroll ,

    Epistemic Frontiers Project,

    Logic and Language, etc.

    Wednesday, March 30, 2005

    What Doctor Does Panniculectomy Covered By Ohip



    In the previous notation, we quoted Article Laws of Thought, published by Keith Devlin in his monthly column Devlin's Angle. Well, just to mention that he, the author an overview of the main features of Stoic logic (created in the heart of the philosophical school Stoics, circa 300 BC), which was the real starting point, together with input about Aristotelian logic (which should however be the systematization of the discipline), the contemporary rationale

    mathematical logic in general and in particular propositional: "[The Stoics] perform a broadening of the Aristotelian syllogistic using five forms of inference or even hypothetical dilemmas, from which they must be able to obtain any meaningful conclusion. While Aristotelian logic was a logic of terms, Stoic logic is a propositional logic. " [Source: 6.3. Stoicism ] The logic propositional logic statements or , really arises both from the contributions of the Stoics to the field of formal logic (as opposed the latter to the logic or material theory of knowledge), and cultivating a logic based on

    propositions (logical statements that associate the values \u200b\u200bof truth and falsity) and not concepts in ( true statements themselves , for example: "All men are mortal, Socrates is a man" needed to infer from them, the demonstration of another, "so therefore Socrates is mortal ") as in the case of Aristotelian logic or syllogistic . A brief look at the main branches of logic has traditionally been considered the remote origin of formal logic or syllogistic logic is classical logic developed by the philosopher Greek Aristotle (384-322 BC), although we have said before to be considered in the development of propositional logic the decisive contribution of Stoic logic. syllogistic logic or Aristotelian , while formalizing common sense, and based on natural language (and therefore not without a certain dose of ambiguity and vagueness) is to establish the truth or falsity of certain philosophical argument. Leaving the field of formalization of logical arguments based on the use of natural language, symbolic or mathematical logic starts with input about the German philosopher GW Leibniz (1646-1717), inspired in turn in the Ars Magna of Ramon Llull (1232-1316) [PDF

    ], trying to focus the logic from the perspective of algebra, while not actually taken into consideration until the nineteenth century, with the help of George Boole (1815-1864), logical and mathematical ; tico British and Boolean logic which gives an algebraic treatment to the propositions of traditional propositional logic. Boole's work in the field of mathematical logic would allow Claude Shannon [1

    ] [ 2] [3 ] developed in the 30's of XX century Bá scheme musician of "Logical machine" digital, initially based on the performance of relays that could take two positions: open or closed (binary logic associated with the operation of an electrical circuit) [

    1] [2
    ]. Symbolic logic using symbols representing the concepts, these being interconnected by operators. In symbolic logic that is developed from the nineteenth century must be distinguished in turn two main branches: propositional logic
    the one hand, and the predicate calculus

    other. propositional logic [ 1] [2 ] is concerned with establishing truth or falsity of one or more propositions, understood as properly formed sentences that can take a true or false value, being connected proposals by operators. For its part, the predicate calculus, also called predicate logic is an extension of propositional logic that takes as basis for calculating the predicate, a function that returns a value

    true or false

    depending on their argument. For example, the knowledge bases of

    logic programming languages \u200b\u200bsuch as Prolog, are based mainly built of predicates that can take into account one or more arguments, or even other predicates to infer processes results in question. Predicate calculus allows for both separation between an object and its possible attributes, and establish relationships between objects and object classes. Moreover, the generalization of predicates is done using variables.

    A step forward in the development of methods of representation knowledge inference based on the use of logical systems, particularly in the development of expert systems on the use that traditionally allowed ló GICA propositional and predicate logic, is the consideration of factors of uncertainty, using probability calculations to that effect from the theory of classical probability, in a first stage of evolution of such systems.

    use of uncertainty factors in becomes essential, while many situations within the observable phenomena in the real world situations support a range much broader than just traditional dichotomy between real

    and false . A union between logic and uncertainty, often referred to as fuzzy logic or fuzzy , and is the basic starting point and essential to the development of systems and applications of Artificial Intelligence (IA). The application of uncertainty factors (chance) in making decisions in situations where information is taken into account hypothetical , by expert systems, is therefore the gateway basic AI. In all cases is a very basic starting point, totally outclassed today by other approaches to addressing the issue of intelligent systems .

    Diy Cobra 148 Peak And Tune

    Stoic Logic Selection Expert System links

    this entry I exit a series of links to articles, reviews, sites and various events, which for whatever reason have earned my attention in recent weeks but of course these listings are representative of anything in particular ... articles, introductions, annotations of "blogs"

    GoogleBrowser ("[...] TouchGraph Java application that creates a relational map a web page based on the results of [...]" via google rivadulla.info

    ). In

    Devlin's Angle, the monthly column Mathematical Association of

    America:

    Claude Shannon ( Castilian translation, March 2001); Laws of Thought (June 1996).

    Functional Programming and XML (XML.com Bijan Parsi in

    , there is also a version of the article translated into French

    ). Besides those mentioned in the article, another example of language based on functional programming paradigm Curry ("A Truly Integrated Functional Logic Language"), but it really is a language that combines the two most relevant paradigms of declarative programming, the aforementioned functional , and the logic programming, resulting in a new amalgamation the programació n

    logical-functional (Functional Logic Programming). User Interface Design Usable: A Quick Guide for Software Developers Free and Open Source

    (text translated by Raúl González Duque Zootropo

    ). Koders

    LFSC'2005 - Symposium on Fuzzy Logic and Soft Computing

    .

    International Conference on Logic Programming (ICLP 2005) are appearing in the log

    Information Retrieval on the Web

    , recommend, as an introduction to the topic, the brief summary on it have long appeared in the weblog
      deceased
    • rivendel.org , entitled Introduction
    • Information Retrieval: Retrieve the Vector Space Model
    • . You can access the text the original location or through Wayback Machine , project recovery tool
    • Internet Archive. (Updated 01/04/2005: Yusef Hassan also echoed this series of excellent articles on See
    • In addition
    • ). I hope to return in the coming weeks a publication rate maso less regular. Thank you all for keeping the interest on this site.

    Monday, January 31, 2005

    How Does The Dome Camera Work

    applied to the extraction information

    One of the articles in the latest issue of Byblos
  • (Electronic Journal of Information Science, n º 20, October-December 2004), is headed title "Test an information extraction system (technical intelligence artificial) in an information center that specializes in plant health. "(the article itself is PDF format
  • ). Written by Ingrid Paz García (Cuba), he describes an expert system applied to the extraction of informació No of scientific articles, whose topics versa in this case plant health. The extraction of information, one of the most important tasks in what has been to be known as mining
  • data [1
  • - PDF -] [
  • 2] [3

    ] [

    4] [5 ] or data mining

    - added additionally, to techniques classic information retrieval, others specific to the field of artificial intelligence to help capture facts (relevant data) from the retrieved documents. By the way, and in conjunction with data mining , there are many experts who do not like the literal translation "mining data ", preferring expressions adjusted original meaning and use of Anglo-Saxon term such as" data mining "or" data discovery ", for example. The expert system described in Arti , ass, used to extract relevant information from scientific articles and technical content, is referred SEISAV (System Information Extraction Plant Protection), and system is based on the CRYSTAL (University of Massachusetts), oriented treatment of texts in English only, so it has been suitably adapted to work with the specific language of the Castilian language, and equipped with more possibilities of use. Induces (build) automatically rules textual analysis from a previous training (method of "automatic training"), but these rules can be constructed manually by an expert familiar in this type of systems, and application-specific domain (method of "Engineering for Knowledge"). The crystal system is a shell (development environment) originally designed to work under MS-DOS, marketed in Europe by Intelligent Environments. However, new environments ( AM for Windows ) relegate CRYSTAL development as a commercial product in the 90's of last century.

    system operation, and some examples of use are clearly explained in the article, quite entertaining, with a level of complexity very affordable even for people without knowledge regarding previous expert systems [1

    2] [3

    ] [ 4] [5 ] [ 6] [7 ], which makes it a highly recommended reading for anyone interested in the implementation of these systems applied to document processing and information management of a significant nature (as opposed to the retrieval and treatment of informació n 'raw').

    Monday, January 3, 2005

    Dont Do Before And After Wax

    reviews of books on logic and demonstration

    Proofs and Logic in Solving Book Reviews are listed all the reviews posted on the web of the Mathematical Association of America (MAA Online ), under section Read This! - The MAA Online book review column on papers whose main theme revolves around logic and demonstration and related disciplines of mathematics, such as the theory to set . To find reviews on other topics in the field of study of mathematics, go to

    subject index. Also noteworthy are the web

    of MAA Online

    , brief monthly columns (on the left drop down menu the portal of entry, are listed under "Publications, Columns "), written by various authors on all sorts of topics related to the broad universe of mathematics, such as Keith Devlin, the oldest of all published under the heading Devlin's Angle ; this author, and the issue of logic and deduction, has something interesting, as are the brief comments The mathematics of human Thought , Laws of Thought and

    Are Mathematicians Turning Soft? , among others. Updated (07/01/2005): In this month's (January 2005), removed Last Doubts about the proof of the Four Color Theorem , K. Devlin realizes new contributions in the demonstration by computerized systems, the problem of Four Colors [1 ] [ 2 ] [3

    , topology, and combinatorics, and gives an overview of the various conjectures raised this issue over time. This problem can be defined, briefly, by the following statement:

    "In a plane or in an area not need more than four colors to color a map so that two neighboring regions, ie, that share a border and not just a point, are not colored the same color " [Source ] John McCarthy, one of the pioneers of Artificial Intelligence, analyzed, Article Coloring Maps and the Kowalski Doctrine

    (1982), a program written in Prolog, and its underlying algorithms, which solved this problem by applying the methodology of logic programming to previous mathematical proofs, reached after the problem statement in the second half of the nineteenth century .