• Àüü
  • ÀüÀÚ/Àü±â
  • Åë½Å
  • ÄÄÇ»ÅÍ
´Ý±â

»çÀÌÆ®¸Ê

Loading..

Please wait....

±¹³» ÇÐȸÁö

Ȩ Ȩ > ¿¬±¸¹®Çå > ±¹³» ÇÐȸÁö > µ¥ÀÌÅͺ£À̽º ¿¬±¸È¸Áö(SIGDB)

µ¥ÀÌÅͺ£À̽º ¿¬±¸È¸Áö(SIGDB)

Current Result Document : 9 / 11 ÀÌÀü°Ç ÀÌÀü°Ç   ´ÙÀ½°Ç ´ÙÀ½°Ç

ÇѱÛÁ¦¸ñ(Korean Title) µ¥ÀÌÅͽºÆ®¸²¿¡¼­ÀÇ Áö¼Ó¼ºÀ» °í·ÁÇÑ °³³ä º¯È­ °ËÃâ ±â¹ý
¿µ¹®Á¦¸ñ(English Title) A Concept Drift Detection Method Considering Persistency Properties in Data Streams
ÀúÀÚ(Author) ±è¸¸¼ö   ÀÓÈ¿»ó   ManSoo Kim   Hyo-Sang Lim  
¿ø¹®¼ö·Ïó(Citation) VOL 34 NO. 02 PP. 0099 ~ 0110 (2018. 08)
Çѱ۳»¿ë
(Korean Abstract)
º» ³í¹®Àº ´ë¿ë·®ÀÇ µ¥ÀÌÅÍ°¡ ½Ç½Ã°£À¸·Î »ðÀԵǴ µ¥ÀÌÅͽºÆ®¸² ȯ°æ¿¡¼­ È¿À²ÀûÀÌ°í Á¤È®ÇÑ °³³ä º¯È­ °ËÃâ ±â¹ýÀ» Á¦¾ÈÇÑ´Ù. º» ¿¬±¸´Â µ¥ÀÌÅͽºÆ®¸² ȯ°æ¿¡¼­ ´ÙÁß Áö¼ö È÷½ºÅä±×·¥(Multiple Exponential Histogram, MEH)À» »ç¿ëÇÏ¿© °³³ä º¯È­¸¦ °ËÃâÇÑ ±âÁ¸ÀÇ ¹æ¹ýÀ» È®ÀåÇÑ´Ù. ±âÁ¸ÀÇ ¹æ¹ýÀº µ¥ÀÌÅͽºÆ®¸²À» ÅëÇØ »ý¼ºµÈ MEH¿¡ µÎ °³ÀÇ ½½¶óÀ̵ù À©µµ¿ì ¹æ¹ýÀ» Àû¿ëÇÑ´Ù. ÀÌ °æ¿ì, µ¥ÀÌÅÍ¿¡¼­ ³ëÀÌÁî(noise)°¡ ¹ß»ýÇϰųª ¿Ï¸¸ÇÑ °³³ä º¯È­°¡ ¹ß»ýÇϸé Á¤»óÀûÀÎ °³³ä º¯È­ °ËÃâÀÌ ¾î·Æ´Ù´Â ´ÜÁ¡ÀÌ ÀÖ´Ù. º» ¿¬±¸´Â °³³ä º¯È­ÀÇ ÀÏ°ü¼º°ú ÇÔ²² Áö¼Ó¼ºÀ» °í·ÁÇÏ°í, ½½¶óÀ̵ù À©µµ¿ì ¹æ½ÄÀ» º¯Çü½ÃÄÑ ÀÌ ¹®Á¦¸¦ ÇØ°áÇÑ´Ù. ½ÇÇèÀ» ÅëÇØ Á¦¾ÈÇÏ´Â ¹æ¹ý¿¡ ´ëÇÑ È¿À²¼º°ú Á¤È®¼ºÀ» º¸ÀδÙ.
¿µ¹®³»¿ë
(English Abstract)
In this paper, we propose an efficient and accurate method for detecting concept drifts in data streams. The proposed method extends our previous work for detecting concept drifts by using Multiple Exponential Histogram(MEH). In the previous work, we have exploited two sliding windows into the MEH which created for input data streams. However, the previous method does not accurately detect concept drifts if there are noises or gradual changes in data streams. We solve these issues with two approaches: 1) considering not only consistency but also persistency properties of concept drifts and 2) adjusting the window sliding method. We also provide the experimental results to show the efficiency and accuracy of the proposed method.
Å°¿öµå(Keyword) Data Stream   Exponential Histogram   Concept Drift   µ¥ÀÌÅͽºÆ®¸²   ´ÙÁß Áö¼ö È÷½ºÅä±×·¥   °³³ä º¯È­ °ËÃâ  
ÆÄÀÏ÷ºÎ PDF ´Ù¿î·Îµå