Á¤º¸Ã³¸®ÇÐȸ ³í¹®Áö ¼ÒÇÁÆ®¿þ¾î ¹× µ¥ÀÌÅÍ °øÇÐ
Current Result Document :
ÇѱÛÁ¦¸ñ(Korean Title) |
°áÇÕµÈ ÆĶó¸ÞÆ®¸¯ È°¼ºÇÔ¼ö¸¦ ÀÌ¿ëÇÑ ÇÕ¼º°ö ½Å°æ¸ÁÀÇ ¼º´É Çâ»ó |
¿µ¹®Á¦¸ñ(English Title) |
Performance Improvement Method of Convolutional Neural Network Using Combined Parametric Activation Functions |
ÀúÀÚ(Author) |
Young Min Ko
Peng Hang Li
Sun Woo Ko
°í¿µ¹Î
À̺ØÇ×
°í¼±¿ì
|
¿ø¹®¼ö·Ïó(Citation) |
VOL 11 NO. 09 PP. 0371 ~ 0380 (2022. 09) |
Çѱ۳»¿ë (Korean Abstract) |
ÇÕ¼º°ö ½Å°æ¸ÁÀº À̹ÌÁö¿Í °°Àº °ÝÀÚ ÇüÅ·Π¹è¿µÈ µ¥ÀÌÅ͸¦ ´Ù·ç´Âµ¥ ³Î¸® »ç¿ëµÇ°í ÀÖ´Â ½Å°æ¸ÁÀÌ´Ù. ÀϹÝÀûÀÎ ÇÕ¼º°ö ½Å°æ¸ÁÀº ÇÕ¼º°öÃþ°ú ¿ÏÀü¿¬°áÃþÀ¸·Î ±¸¼ºµÇ¸ç °¢ ÃþÀº ºñ¼±ÇüÈ°¼ºÇÔ¼ö¸¦ Æ÷ÇÔÇÏ°í ÀÖ´Ù. º» ³í¹®Àº ÇÕ¼º°ö ½Å°æ¸ÁÀÇ ¼º´ÉÀ» Çâ»ó½ÃÅ°±â À§ÇØ °áÇÕµÈ ÆĶó¸ÞÆ®¸¯ È°¼ºÇÔ¼ö¸¦ Á¦¾ÈÇÑ´Ù. °áÇÕµÈ ÆĶó¸ÞÆ®¸¯ È°¼ºÇÔ¼ö´Â È°¼ºÇÔ¼öÀÇ Å©±â¿Í À§Ä¡¸¦ º¯È¯½ÃÅ°´Â ÆĶó¹ÌÅ͸¦ Àû¿ëÇÑ ÆĶó¸ÞÆ®¸¯ È°¼ºÇÔ¼öµéÀ» ¿©·¯ ¹ø ´õÇÏ¿© ¸¸µé¾îÁø´Ù. ¿©·¯ °³ÀÇ Å©±â, À§Ä¡¸¦ º¯È¯ÇÏ´Â ÆĶó¹ÌÅÍ¿¡ µû¶ó ´Ù¾çÇÑ ºñ¼±Çü°£°ÝÀ» ¸¸µé ¼ö ÀÖÀ¸¸ç, ÆĶó¹ÌÅÍ´Â ÁÖ¾îÁø ÀԷµ¥ÀÌÅÍ¿¡ ÀÇÇØ °è»êµÈ ¼Õ½ÇÇÔ¼ö¸¦ ÃÖ¼ÒÈÇÏ´Â ¹æÇâÀ¸·Î ÇнÀÇÒ ¼ö ÀÖ´Ù. °áÇÕµÈ ÆĶó¸ÞÆ®¸¯ È°¼ºÇÔ¼ö¸¦ »ç¿ëÇÑ ÇÕ¼º°ö ½Å°æ¸ÁÀÇ ¼º´ÉÀ» MNIST, Fashion MNIST, CIFAR10 ±×¸®°í CIFAR100 ºÐ·ù¹®Á¦¿¡ ´ëÇØ ½ÇÇèÇÑ °á°ú, ´Ù¸¥ È°¼ºÇÔ¼öµéº¸´Ù ¿ì¼öÇÑ ¼º´ÉÀ» °¡ÁüÀ» È®ÀÎÇÏ¿´´Ù.
|
¿µ¹®³»¿ë (English Abstract) |
Convolutional neural networks are widely used to manipulate data arranged in a grid, such as images. A general convolutional neural network consists of a convolutional layers and a fully connected layers, and each layer contains a nonlinear activation functions. This paper proposes a combined parametric activation function to improve the performance of convolutional neural networks. The combined parametric activation function is created by adding the parametric activation functions to which parameters that convert the scale and location of the activation function are applied. Various nonlinear intervals can be created according to parameters that convert multiple scales and locations, and parameters can be learned in the direction of minimizing the loss function calculated by the given input data. As a result of testing the performance of the convolutional neural network using the combined parametric activation function on the MNIST, Fashion MNIST, CIFAR10 and CIFAR100 classification problems, it was confirmed that it had better performance than other activation functions. |
Å°¿öµå(Keyword) |
Convolutional Neural Network
Nonlinear Activation Function
Combined Parametric Activation Function
Loss function
ÇÕ¼º°ö ½Å°æ¸Á
ºñ¼±ÇüÈ°¼ºÇÔ¼ö
°áÇÕµÈ ÆĶó¸ÞÆ®¸¯ È°¼ºÇÔ¼ö
¼Õ½ÇÇÔ¼ö
|
ÆÄÀÏ÷ºÎ |
PDF ´Ù¿î·Îµå
|