ãDeep Learning Implementationãã§ã°ã°ãã¨ãããã«åºã¦ãããã¨ã§, ãã¾ã«è©±é¡ã«ãªãã©ã¤ãã©ãª Theano ã«ã¤ãã¦ç°¡åã«èª¿ã¹ãã®ã§ç´¹ä»ãã. ééãããã£ããæ¯éæãã¦æ¬²ãã. ãªã, Deep Learningã«é¢ãã解説ã¯ä»åã¯è¡ããªã. Theanoã¨ã¯Pythonã«ãããæ°å¤è¨ç®ç¨ã®ã©ã¤ãã©ãªã§ãã. è¡åæ¼ç®ãªã©ãè¡ãé¢æ°ãæä¾ãã¦ãã, numpy/scipyã®ä»£æ¿ããã±ã¼ã¸ã¨æãã°ããã ãã. 大ããªç¹å¾´ã¯ å®è¡æã«ãããC++ã³ã¼ãã®çæã¨ã³ã³ãã¤ã« GPUãµãã¼ã 解æçãªå¾®åã®ãµãã¼ã (x^2ã®å¾®å=2xã¨ããæä½ãèªåã§ãã£ã¦ããã) ã®3ã¤ã§ãã. ä¸é¨ã¯å¾ã«è©³ç´°ã説æãã. ããã«ãã£ã¦Multilayer Perceptronãå®è¡ãããã³ããã¼ã¯ã§ã¯, Theanoãnumpyãã1.8å, Matlabãã1.6åéãã¨ãã
{{#tags}}- {{label}}
{{/tags}}