{"id":1073,"date":"2020-07-31T23:30:47","date_gmt":"2020-08-01T03:30:47","guid":{"rendered":"http:\/\/aristotle2digital.blogwyrm.com\/?p=1073"},"modified":"2020-07-31T19:00:25","modified_gmt":"2020-07-31T23:00:25","slug":"machine-classification-part-1-overview","status":"publish","type":"post","link":"https:\/\/aristotle2digital.blogwyrm.com\/?p=1073","title":{"rendered":"Machine Classification: Part 1 \u2013 Overview"},"content":{"rendered":"\n

Classification is one of the key components of the modern AI\ntoolkit in which a machine learning (ML) algorithm attempts to mimic the human\nability to distinguish and categorize.  The\nidea is that the algorithm, when confronted with a new instance of an object,\nis to statistically determine the class or category into which this object best\nfits.   <\/p>\n\n\n\n

Classification is one of those human activities that is deceptively simple.\u00a0 For example, for decades people thought that the horseshoe crab<\/a> was related to the crustaceans<\/a> because it could be found in the ocean.\u00a0 As biology progressed, it became clearer that the horseshoe crab wasn\u2019t in the same class as the crustaceans and that it has more in common with arachnids<\/a>, making it the \u2018spider of the seas\u2019.\u00a0 <\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

The ability of an expert to determine whether an object belongs in one category is also a subtle affair that is often as much art as it is science, as the following excerpt from Miss Marple\u2019s speech in A Christmas Tragedy<\/em> by Agatha Christie nicely describes:<\/p>\n\n\n\n

\n

It\u2019s really a matter of practice and experience.  An Egyptologist, so I\u2019ve heard, if you show\nhim one of those curious little beetles, can tell you by the look and feel of\nthe thing what date B.C. it is, or if it\u2019s a Birmingham imitation.  And he can\u2019t always give a definite rule for\ndoing so.  He just knows<\/em>.  His life has been spent handling such things.<\/div>\n\n\n\n

Christie makes several important points in that brief passage.\u00a0 First, there is the matter of \u2018practice and experience\u2019.\u00a0 This translates in the domain of machine learning to training.\u00a0 Second, she speaks of the Egyptologist \u2018handling\u2019 the beetle and judging by the \u2018look and the feel of the thing\u2019.\u00a0 This requirement corresponds to having a set of percepts about the object, a point that is, arguably, the trickiest.\u00a0 The third point she raises is that the expert can classify the age of the object (\u2018what date B.C\u2019) or can spot a counterfeit.\u00a0 Of course, this is the point of the ML algorithm in the first place, to be able to judge expertly a new object.\u00a0 The fourth and final point is that the expert can\u2019t always give a definite rule explaining how he judged.\u00a0 There is no direct translation of this rule into the domain of machine learning, but more on that point below.<\/p>\n\n\n\n

In thinking about an expert (Egyptologist or otherwise), we need to recognize that what makes him an expert is that he is more often right than wrong.\u00a0 The context for the previous excerpt is the argument that Miss Marple, a spinster sleuth of uncertain age, makes about how \u2018superfluous women\u2019 (such as herself) who engage in \u2018tittle tattle\u2019 are \u2018nine times out of ten\u2019 correct, and \u2018[t]hat\u2019s really just what makes people so annoyed about it\u2019.\u00a0 So, we can\u2019t expect our machine learning algorithm to be able to be 100% accurate, since no expert ever is; we can only hope that it is \u2018accurate enough\u2019.\u00a0 <\/p>\n\n\n\n

It is also very likely that the algorithm will never be as\naccurate as a human expert for the following reason.  In philosophical terms, machine\nclassification overlaps the first and second Acts of the Mind<\/a> (the Act\nof Understanding and the Act of Judgement) without, necessarily, being fully\ndeveloped in either.   <\/p>\n\n\n\n

For humans, the first act involves apprehending the percepts provided (\u2018handling\u2019 the object to get an idea of its \u2018look and feel\u2019).\u00a0 A baby is born with the ability to process his perceptions; to make some sort of comprehension of the sensory input from the five senses. In the second act, the person abstracts universals (or what, at least functionally, passes as such) from those sensory experiences to be able to understand \u2018redness\u2019 or \u2018roundness\u2019 or the being-qua-being of any other form.\u00a0 These universals allow the human to then classify and sub-classify the objects in the surrounding world.\u00a0 <\/p>\n\n\n\n

In contrast, the machine is taught about only a small subset of possible percepts (typically digital data representing an image or a time series).\u00a0 Currently, no machine can expand or contract its attention when it realizes it needs to know more or is being blasted with too much information.\u00a0 In addition, it only knows the categories that are used to train it.\u00a0 <\/p>\n\n\n\n

The human has a decided advantage in that he can expand or contract the number of attributes used in the classification on the fly (e.g., concentrating only the weight and the texture first and then adding in color and style as needed later) and the human can invent new attributes at need (e.g. suddenly noticing that the size matters).\u00a0 The machine has only two advantages: raw speed and the ability to handle an arbitrarily large number of attributes (although the number must be fixed for each situation).\u00a0 As a result, the machine’s ability to classify is entirely based on some statistical or probabilistic measure.\u00a0 The human\u2019s ability to classify is surely rooted in probability as well, but what if anything else is going on is, at this time, anybody\u2019s guess.<\/p>\n\n\n\n

To be more concrete, consider the problem of spam\nemails.  Determining whether a given email\nis spam or not is a good example of a classification problem that illustrates some\nof the advantages and disadvantages on both sides.  The human can actually read the content of an\nemail, comprehend the meanings, and judge the context (which may require consideration\nof different attributes compared to the previous email) before deciding whether\nthe message is good or bad.  However, the\nhuman can only read a limited number of emails each day and is prone to getting\nbored or tired and making mistakes.  The\nmachine can make sense of large amount of the associated network data, be it IP\naddresses, message size, number of hops and so on \u2013 data that would make little\nor no sense to an overwhelming number of most humans.  In addition, the machine can analyze a vast number\nof messages in the time it takes the human to read one.  <\/p>\n\n\n\n

Over the coming months, this column will look at some of the more popular ML techniques for classifying data<\/a> and compare the pros and cons of each technique.\u00a0 Some of the metrics for the comparison will be the difficulty of assembling a training set (the data that gives the required \u2018practice and experience\u2019), whether the data need to be pre-labeled into classes (e.g., a real scarab or a Birmingham imitation), or whether we can allow the algorithm to find the possible classes based on how the data cluster, the accuracy of the method compared to truth, and the application domains that experts use.\u00a0 In the end, we will have essentially a classification of classification algorithms. <\/p>\n","protected":false},"excerpt":{"rendered":"

Classification is one of the key components of the modern AI toolkit in which a machine learning (ML) algorithm attempts to mimic the human ability to distinguish and categorize.  The… Read more ><\/a><\/p>\n","protected":false},"author":2,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[],"class_list":["post-1073","post","type-post","status-publish","format-standard","hentry","category-uncategorized"],"_links":{"self":[{"href":"https:\/\/aristotle2digital.blogwyrm.com\/index.php?rest_route=\/wp\/v2\/posts\/1073","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/aristotle2digital.blogwyrm.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/aristotle2digital.blogwyrm.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/aristotle2digital.blogwyrm.com\/index.php?rest_route=\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/aristotle2digital.blogwyrm.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=1073"}],"version-history":[{"count":0,"href":"https:\/\/aristotle2digital.blogwyrm.com\/index.php?rest_route=\/wp\/v2\/posts\/1073\/revisions"}],"wp:attachment":[{"href":"https:\/\/aristotle2digital.blogwyrm.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=1073"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/aristotle2digital.blogwyrm.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=1073"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/aristotle2digital.blogwyrm.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=1073"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}