Scribd Logo
Upload

Welcome to Scribd!

  • 18 views

    Aprendizaje de Maquinas - Lectura FINAL

    Uploaded by

    lina cardozo
    Machine learning is a method of data analysis that automates analytical model building. It allows systems to learn from data, identifying patterns and making decisions with minimal human involvement. There are three main types of machine learning: supervised learning where the system learns from labeled examples; unsupervised learning which finds hidden patterns in unlabeled data; and reinforcement learning where an agent learns to maximize rewards. Machine learning has increased the capabilities of computer systems by allowing them to perform complex tasks like visual recognition far better than traditional programming. However, human performance still exceeds machines for more general visual recognition and other complex tasks.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd

    Aprendizaje de Maquinas - Lectura FINAL

    Uploaded by

    lina cardozo
    18 views2 pages
    Machine learning is a method of data analysis that automates analytical model building. It allows systems to learn from data, identifying patterns and making decisions with minimal human involvement. There are three main types of machine learning: supervised learning where the system learns from labeled examples; unsupervised learning which finds hidden patterns in unlabeled data; and reinforcement learning where an agent learns to maximize rewards. Machine learning has increased the capabilities of computer systems by allowing them to perform complex tasks like visual recognition far better than traditional programming. However, human performance still exceeds machines for more general visual recognition and other complex tasks.

    Original Title

    Aprendizaje de maquinas_Lectura FINAL

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Machine learning is a method of data analysis that automates analytical model building. It allows systems to learn from data, identifying patterns and making decisions with minimal human involvement. There are three main types of machine learning: supervised learning where the system learns from labeled examples; unsupervised learning which finds hidden patterns in unlabeled data; and reinforcement learning where an agent learns to maximize rewards. Machine learning has increased the capabilities of computer systems by allowing them to perform complex tasks like visual recognition far better than traditional programming. However, human performance still exceeds machines for more general visual recognition and other complex tasks.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    18 views2 pages

    Aprendizaje de Maquinas - Lectura FINAL

    Uploaded by

    lina cardozo
    Machine learning is a method of data analysis that automates analytical model building. It allows systems to learn from data, identifying patterns and making decisions with minimal human involvement. There are three main types of machine learning: supervised learning where the system learns from labeled examples; unsupervised learning which finds hidden patterns in unlabeled data; and reinforcement learning where an agent learns to maximize rewards. Machine learning has increased the capabilities of computer systems by allowing them to perform complex tasks like visual recognition far better than traditional programming. However, human performance still exceeds machines for more general visual recognition and other complex tasks.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    of 2

    CHAPTER ONE

    1.3 What is machine learning?


    Machine earning
    with the term AI, the abi ity to earn rom In one image
    computers to earn direct y rom examp es data increases the number and comp exity
    the accuracy o
    approaches to programming re y on hard- can undertake, in comparison to traditiona
    machine earning
    coded ru es, which set out how to so ve a programming methods. Machine earning
    prob em, step-by-step. In contrast, machine can carry out tasks o such comp exity that has increased
    earning systems are set a task, and given a rom 72% in
    programs based on step-by-step processes
    2010, to 96% in
    how this task can be achieved or rom which created by humans. The earning e ement
    to detect patterns. The system then earns a so creates systems which can be adaptive, 2015, surpassing
    how best to achieve the desired output. It can and continue to improve the accuracy o their human accuracy
    9
    . at  this  task.
    supports inte igent systems, which are ab e to

    computer science, statistics, and data science.

    In some speci ic areas or tasks, machine process data in a way that can detect and

    eve o per ormance than peop e. For other or  make decisions.
    tasks, human per ormance remains much
    better than that o machine earning systems.
    For examp e, recent advances in image
    recognition have made these systems more
    accurate than ever be ore. In one image

    96% in 2015, surpassing human accuracy at


    this task 8. However, human- eve per ormance
    at visua recognition in more genera terms
    remains considerab y higher than these
    systems can achieve.

    8.

    9.
    Cambridge, UK: Cambridge University Press.

    MACHINE LEARNING: THE POWER AND PROMISE OF COMPUTERS THAT LEARN BY EXAMPLE 19
    CHAPTER ONE

    There are three key branches o machine


    earning: on ine earning systems:

    • In supervised machine earning, a system •

    one or more groups, such as ‘app es’ or


    ‘oranges’. The system earns how this data
    – known as training data – is structured, and setting, tested, and then dep oyed using
    uses this to predict the categories o new –
    or  ‘test’ – data. methods. This approach is more common in

    • Unsupervised earning is earning without


    today, because it gives an opportunity or

    that make data points more or ess simi ar to


    system interacts with any user.

    and assigning data to these  c usters. •

    • Rein orcement earning ocuses on


    but the earning a gorithms continue to be

    unsupervised and supervised earning. In


    This means that the per ormance o the
    system ‘in the wi d’ can continue to improve in
    agent10 interacts with its environment, and is
    given a reward unction that it tries to optimise,
    means that there is no opportunity or human

    the mode , be ore users are exposed to these.


    earn the consequences o its decisions, such
    as which moves were important in winning
    a game, and to use this earning to ind
    strategies that maximise its  rewards.
    eedback on the system’s accuracy.

    10.

    20 MACHINE LEARNING: THE POWER AND PROMISE OF COMPUTERS THAT LEARN BY EXAMPLE

    You might also like