{"id":2308,"date":"2017-02-27T20:36:48","date_gmt":"2017-02-28T04:36:48","guid":{"rendered":"http:\/\/blog.chrisrivard.com\/?p=2308"},"modified":"2017-02-27T20:36:48","modified_gmt":"2017-02-28T04:36:48","slug":"some-ml-notes","status":"publish","type":"post","link":"https:\/\/www.chrisrivard.com\/blog\/some-ml-notes\/","title":{"rendered":"Some ML Notes"},"content":{"rendered":"

Supervised, Unsupervised, Semi-supervised, Reinforcement learning<\/h3>\n
    \n
  1. Supervised<\/li>\n

    Supervised learning procedures are used in problems for which we
    \ncan provide the system with example inputs as well as their corre\u2010
    \nsponding outputs and wish to induce an implicit approximation of
    \nthe rules or function that governs these correlations.<\/p>\n

    The kinds of problems that can be addressed by supervised learning procedures are generally divided into two categories: classification\u00a0and regression problems.<\/p>\n

    In a classification problem, the outputs relate to a set of discrete categories.
    \nFor example, we may have an image of a handwritten character and
    \nwish to determine which of 26 possible letters it represents. In a
    \nregression problem, the outputs relate to a real-valued number. For
    \nexample, based on a set of financial metrics and past performance
    \ndata, we may try to guess the future price of a particular stock.<\/p>\n

  2. Unsupervised<\/li>\n

    Unsupervised learning procedures do not require a set of known out\u2010
    \nputs. Instead, the machine is tasked with finding internal patterns
    \nwithin the training examples. Procedures of this kind are \u201cunsuper\u2010
    \nvised\u201d in the sense that we do not explicitly indicate what the system
    \nshould learn about. Instead, we provide a set of training examples
    \nthat we believe contains internal patterns and leave it to the system
    \nto discover those patterns on its own.<\/p>\n

    In general, unsupervised learning can provide assistance in our efforts to understand extremely complex systems whose internal patterns may be too
    \ncomplex for humans to discover on their own. Unsupervised learn\u2010
    \ning can also be used to produce generative models…<\/p>\n

  3. Semi-supervised<\/li>\n

    Semi-supervised learning procedures use the automatic feature dis\u2010
    \ncovery capabilities of unsupervised learning systems to improve the
    \nquality of predictions in a supervised learning problem. Instead of
    \ntrying to correlate raw input data with the known outputs, the raw
    \ninputs are first interpreted by an unsupervised system. The unsuper\u2010
    \nvised system tries to discover internal patterns within the raw input
    \ndata, removing some of the noise and helping to bring forward the
    \nmost important or indicative features of the data. These distilled ver\u2010
    \nsions of the data are then handed over to a supervised learning
    \nmodel, which correlates the distilled inputs with their correspond\u2010
    \ning outputs in order to produce a predictive model whose accuracy
    \nis generally far greater than that of a purely supervised learning system.<\/p>\n

  4. Reinforcement learning<\/li>\n

    Reinforcement learning procedures use rewards and punishments to
    \nshape the behavior of a system with respect to one or several specific
    \ngoals. Unlike supervised and unsupervised learning systems, rein\u2010
    \nforcement learning systems are not generally trained on an existent
    \ndataset and instead learn primarily from the feedback they gather
    \nthrough performing actions and observing the consequences.<\/p>\n<\/ol>\n

    From Machine Learning for Designers<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"

    Supervised, Unsupervised, Semi-supervised, Reinforcement learning Supervised Supervised learning procedures are used in problems for which we can provide the system with example inputs as well as their corre\u2010 sponding outputs and wish to induce an implicit approximation of the rules or function that governs these correlations. The kinds of problems that can be addressed by […]<\/p>\n","protected":false},"author":2,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[4],"tags":[],"class_list":["post-2308","post","type-post","status-publish","format-standard","hentry","category-sartor-resartus"],"_links":{"self":[{"href":"https:\/\/www.chrisrivard.com\/blog\/wp-json\/wp\/v2\/posts\/2308","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.chrisrivard.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.chrisrivard.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.chrisrivard.com\/blog\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.chrisrivard.com\/blog\/wp-json\/wp\/v2\/comments?post=2308"}],"version-history":[{"count":0,"href":"https:\/\/www.chrisrivard.com\/blog\/wp-json\/wp\/v2\/posts\/2308\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.chrisrivard.com\/blog\/wp-json\/wp\/v2\/media?parent=2308"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.chrisrivard.com\/blog\/wp-json\/wp\/v2\/categories?post=2308"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.chrisrivard.com\/blog\/wp-json\/wp\/v2\/tags?post=2308"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}