WHAT IS MA­CHINE LEARN­ING?

Focus-Science and Technology - - In Focus -

Ma­chine learn­ing is a branch QH CTVKƂEKCN KPVGNNKIGPEG VJCV bor­rows heav­ily from other ƂGNFU UWEJ CU UVCVKUVKEU math­e­mat­ics, physics and the­o­ret­i­cal com­puter science. Ma­chine-learn­ing al­go­rithms en­able com­put­ers to per­form tasks with­out hu­man in­ter­ven­tion. With this tech­nique, ma­chines can be taught to make their own pro­grams. Ev­ery time you ask your vir­tual as­sis­tant Alexa QT 5KTK VQ RGTHQTO C URGEKƂE task, that’s ma­chine learn­ing in ac­tion.

More com­plex uses of ma­chine learn­ing in­clude KOCIG ENCUUKƂECVKQP (QT ex­am­ple, Adobe has de­vel­oped a fea­ture called Smart Tags for its pro­gram Ex­pe­ri­ence Man­ager. It can add tags to im­ages au­to­mat­i­cally. For ex­am­ple, to an im­age of the Wash­ing­ton Me­mo­rial, it would slap tags such as ‘ar­chi­tec­ture’, ‘me­mo­rial’, and ‘Wash­ing­ton Me­mo­rial’. If some­one searched for ar­chi­tec­ture im­ages on their sys­tem, the Wash­ing­ton Me­mo­rial photo would show up. AI art tools de­vel­oped by Mario Klinge­mann, Dan Ven­tura and Si­mon Colton all make use of ma­chine learn­ing.

Newspapers in English

Newspapers from UK

© PressReader. All rights reserved.