Scribd Logo
Upload

Welcome to Scribd!

  • Artificial Neural Networks - Utec Ai

    Uploaded by

    Bregy Esteban Malpartida Ramos
    42 views13 pages
    The document summarizes the early history and developments of artificial neural networks, beginning with McCulloch and Pitts' 1943 model of how neurons might work using electrical circuits. It then discusses Donald Hebb's 1949 work on neural plasticity, Rochester's failed 1950 attempt to simulate a neural network, and Frank Rosenblatt's development of the Perceptron in 1958. Later sections cover ADALINE and MADALINE networks developed by Widrow and Hoff in 1959, their introduction of gradient descent learning in 1962, and John Hopfield's influential 1982 work on bidirectional neural networks.

    Original Description:

    We're a university club with deep love to AI, and this is our first course about Machine Learning.

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Report this Document
    The document summarizes the early history and developments of artificial neural networks, beginning with McCulloch and Pitts' 1943 model of how neurons might work using electrical circuits. It then discusses Donald Hebb's 1949 work on neural plasticity, Rochester's failed 1950 attempt to simulate a neural network, and Frank Rosenblatt's development of the Perceptron in 1958. Later sections cover ADALINE and MADALINE networks developed by Widrow and Hoff in 1959, their introduction of gradient descent learning in 1962, and John Hopfield's influential 1982 work on bidirectional neural networks.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Flag for inappropriate content
    42 views13 pages

    Artificial Neural Networks - Utec Ai

    Uploaded by

    Bregy Esteban Malpartida Ramos
    The document summarizes the early history and developments of artificial neural networks, beginning with McCulloch and Pitts' 1943 model of how neurons might work using electrical circuits. It then discusses Donald Hebb's 1949 work on neural plasticity, Rochester's failed 1950 attempt to simulate a neural network, and Frank Rosenblatt's development of the Perceptron in 1958. Later sections cover ADALINE and MADALINE networks developed by Widrow and Hoff in 1959, their introduction of gradient descent learning in 1962, and John Hopfield's influential 1982 work on bidirectional neural networks.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Flag for inappropriate content
    of 13

    We can only see a short distance ahead, but we can see plenty there that needs to be done.

    ― Alan Turing
    ARTIFICIAL NEURAL NETWORKS

    1943. First biological and artificial approximation

    Neurophysiologist Warren McCulloch and


    mathematician Walter Pitts wrote a paper on how
    neurons might work. In order to describe how
    neurons in the brain might work, they modeled a
    simple neural network using electrical circuits.

    Hayman, S. (1999). The McCulloch-Pitts model.


    6. 4438 - 4439 vol.6.
    10.1109/IJCNN.1999.830886.

    http://www.mind.ilstu.edu/curriculum/modOverview.php?modGUI=21
    2
    ARTIFICIAL NEURAL NETWORKS

    1943. Logic Units with Threshold

    Using McCulloch-Pitts model we can says our first


    artificial neuron approximation also called LUT. The
    inputs could be either a zero or a one.

    http://aishack.in/tutorials/artificial-neurons-mccullochpitts-model/
    ARTIFICIAL NEURAL NETWORKS

    1949. Donald Hebb wrote The Organization of Behavior

    Donald Hebb wrote The Organization of Behavior, a


    work which pointed out the fact that neural pathways
    are strengthened each time they are used, a concept
    fundamentally essential to the ways in which humans
    learn. If two nerves fire at the same time, he argued, the
    connection between them is enhanced.

    http://neuronaldynamics.epfl.ch/online/Ch19.S1.htm
    l
    ARTIFICIAL NEURAL NETWORKS

    ~ 1950

    As computers became more advanced in the 1950's, it


    was finally possible to simulate a hypothetical neural
    network. The first step towards this was made by
    Nathanial Rochester from the IBM research laboratories.
    Unfortunately for him, the first attempt to do so failed.

    Around this time, Frank Rosenblatt, a psychologist at


    Cornell, was working on understanding the
    comparatively simpler decision systems present in the
    eye of a fly, which underlie and determine its flee
    response. In an attempt to understand and quantify this
    https://medium.com/@Jaconda/a-concise-history-of-neural-net process, he proposed the idea of a Perceptron in 1958
    works-2070655d3fec
    ARTIFICIAL NEURAL NETWORKS

    ~ 1950

    First known implementation of a Mark I Perceptron.


    The machine was connected to a camera that used
    20×20 cadmium sulfide photocells to produce a
    400-pixel image. The main visible feature is a
    patchboard that allowed experimentation with
    different combinations of input features. To the right
    of that are arrays of potentiometers that
    implemented the adaptive weights.[wiki]
    ARTIFICIAL NEURAL NETWORKS

    1959. ADELINE and MADELINE networks

    Bernard Widrow and Marcian Hoff of Stanford


    developed models called "ADALINE" and "MADALINE."
    ADALINE was developed to recognize binary patterns so
    that if it was reading streaming bits from a phone line, it
    could predict the next bit. MADALINE was the first neural
    network applied to a real world problem, using an
    adaptive filter that eliminates echoes on phone lines.
    While the system is as ancient as air traffic control
    systems, like air traffic control systems, it is still in
    commercial use.
    https://en.wikipedia.org/wiki/ADALINE
    ARTIFICIAL NEURAL NETWORKS

    1962. Widrow & Hoff and first learning processes.

    Magic starts here!. Gradient descent method

    https://www.ibu.edu.ba/assets/userfiles/it/2012/eee-ANN-4.pd
    f
    ARTIFICIAL NEURAL NETWORKS

    1962. Widrow & Hoff and first learning processes

    Useful equations for train a simple perceptron model.


    ARTIFICIAL NEURAL NETWORKS

    1962. Widrow & Hoff and first learning processes


    ARTIFICIAL NEURAL NETWORKS

    1982. ADELINE and MADELINE networks

    In 1982, interest in the field was renewed. John Hopfield


    of Caltech presented a paper to the National Academy
    of Sciences. His approach was to create more useful
    machines by using bidirectional lines. Previously, the
    connections between neurons was only one way.

    https://en.wikipedia.org/wiki/Hopfield_networ
    k
    ARTIFICIAL NEURAL NETWORKS

    1982. ADELINE and MADELINE networks

    https://en.wikipedia.org/wiki/Hopfield_networ
    k

    You might also like