BEGIN:VCALENDAR VERSION:2.0 PRODID:IEEE vTools.Events//EN CALSCALE:GREGORIAN BEGIN:VTIMEZONE TZID:Canada/Eastern BEGIN:DAYLIGHT DTSTART:20220313T030000 TZOFFSETFROM:-0500 TZOFFSETTO:-0400 RRULE:FREQ=YEARLY;BYDAY=2SU;BYMONTH=3 TZNAME:EDT END:DAYLIGHT BEGIN:STANDARD DTSTART:20211107T010000 TZOFFSETFROM:-0400 TZOFFSETTO:-0500 RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=11 TZNAME:EST END:STANDARD END:VTIMEZONE BEGIN:VEVENT DTSTAMP:20220328T183909Z UID:5C39333A-90E8-4F16-AEB6-1528E2D52D5A DTSTART;TZID=Canada/Eastern:20211109T170000 DTEND;TZID=Canada/Eastern:20211109T183000 DESCRIPTION:--- Prerequisites ---\n\nYou do not need to have attended the e arlier talks. If you know zero math and zero machine learning\, then this talk is for you. Jeff will do his best to explain fairly hard mathematics to you. If you know a bunch of math and/or a bunch machine learning\, then these talks are for you. Jeff tries to spin the ideas in new ways.\n\n--- Longer Abstract ---\n\nAn input data item\, eg a image of a cat\, is just a large tuple of real values. As such it can be thought as a point in som e high dimensional vector space. Whether the image is of a cat or a dog pa rtitions this vector space into regions. Classifying your image amounts to knowing which region the corresponding point is in. The dot product of tw o vectors tell us: whether our data scaled by coefficients meets a thresho ld\; how much two lists of properties correlate\; the cosine of the angle between to directions\; and which side of a hyperplane your points is on. A novice reading a machine learning paper might not get that many of the s ymbols are not real numbers but are matrices. Hence the product of two suc h symbols is matrix multiplication. Computing the output of your current n eural network on each of your training data items amounts to an alternatio n of such a matrix multiplications and of some non-linear rounding of your numbers to be closer to being 0-1 valued. Similarly\, back propagation co mputes the direction of steepest decent using a similar alternation\, exce pt backwards. The matrix way of thinking about a neural network also helps us understand how a neural network effectively performs a sequence linear and non-linear transformations changing the representation of our input u ntil the representation is one for which the answer can be determined base d which side of a hyperplane your point is on. Though people say that it i s "obvious"\, it was never clear to me which direction to head to get the steepest decent. Slides Covered: http://www.eecs.yorku.ca/~jeff/courses/ma chine-learning\n\n/Machine_Learning_Made_Easy.pptx\n\n- Linear Regression\ , Linear Separator\n\n- Neural Networks\n\n- Abstract Representations\n\n- Matrix Multiplication\n\n- Example\n\n- Vectors\n\n- Back Propagation\n\n - Sigmoid\n\nSpeaker(s): Prof. Jeff Edmonds\, \n\nVirtual: https://events. vtools.ieee.org/m/287446 LOCATION:Virtual: https://events.vtools.ieee.org/m/287446 ORGANIZER:ayda.naserialiabadi@ryerson.ca SEQUENCE:1 SUMMARY:Algebra Review: How does one best think about all of these numbers URL;VALUE=URI:https://events.vtools.ieee.org/m/287446 X-ALT-DESC:Description: <br /><p>--- Prerequisites ---</p>\n<p>You do not n eed to have attended the earlier talks. If you know zero math and zero mac hine learning\, then this talk is for you. Jeff will do his best to explai n fairly hard mathematics to you. If you know a bunch of math and/or a bun ch machine learning\, then these talks are for you. Jeff tries to spin the ideas in new ways.</p>\n<p>--- Longer Abstract ---</p>\n<p>An input data item\, eg a image of a cat\, is just a large tuple of real values. As such it can be thought as a point in some high dimensional vector space. Wheth er the image is of a cat or a dog partitions this vector space into region s. Classifying your image amounts to knowing which region the correspondin g point is in. The dot product of two vectors tell us: whether our data sc aled by coefficients meets a threshold\; how much two lists of properties correlate\; the cosine of the angle between to directions\; and which side of a hyperplane your points is on. A novice reading a machine learning pa per might not get that many of the symbols are not real numbers but are ma trices. Hence the product of two such symbols is matrix multiplication. Co mputing the output of your current neural network on each of your training data items amounts to an alternation of such a matrix multiplications and of some non-linear rounding of your numbers to be closer to being 0-1 val ued. Similarly\, back propagation computes the direction of steepest decen t using a similar alternation\, except backwards. The matrix way of thinki ng about a neural network also helps us understand how a neural network ef fectively performs a sequence linear and non-linear transformations changi ng the representation of our input until the representation is one for whi ch the answer can be determined based which side of a hyperplane your poin t is on. Though people say that it is "obvious"\, it was never clear to me which direction to head to get the steepest decent. Slides Covered: http: //www.eecs.yorku.ca/~jeff/courses/machine-learning</p>\n<p>/Machine_Learni ng_Made_Easy.pptx</p>\n<p>- Linear Regression\, Linear Separator</p>\n<p>- Neural Networks</p>\n<p>- Abstract Representations</p>\n<p>- Matrix Multi plication</p>\n<p>- Example</p>\n<p>- Vectors</p>\n<p>- Back Propagation</ p>\n<p>- Sigmoid</p> END:VEVENT END:VCALENDAR