Çѱ¹ÀÎÅͳÝÁ¤º¸ÇÐȸ ³í¹®Áö
Current Result Document :
ÇѱÛÁ¦¸ñ(Korean Title) |
È®ÀåµÈ °ÈÇнÀ ½Ã½ºÅÛÀÇ Á¤Çü¸ðµ¨ |
¿µ¹®Á¦¸ñ(English Title) |
Formal Model of Extended Reinforcement Learning (E-RL) System |
ÀúÀÚ(Author) |
À±Áö¿µ
±èµ¿¿í
½Å°ÇÀ±
±è»ó¼ö
ÇÑ¸í¹¬
Jiyoung Yun
Dong-Wook Kim
Gun-Yoon Shin
Sang-Soo Kim
Myung-Mook Han
Àüµµ¿µ
¼Û¸íÈ£
±è¼öµ¿
Do Yeong Jeon
Myeong Ho Song
Soo Dong Kim
|
¿ø¹®¼ö·Ïó(Citation) |
VOL 22 NO. 04 PP. 0013 ~ 0028 (2021. 08) |
Çѱ۳»¿ë (Korean Abstract) |
°ÈÇнÀÀº ÇÑ È¯°æ¿¡¼ ¿¡ÀÌÀüÆ®°¡ Á¤Ã¥¿¡ µû¶ó ¾×¼ÇÀ» ÃëÇÏ°í º¸»ó ÇÔ¼ö¸¦ ÅëÇØ ¾×¼ÇÀ» Æò°¡ ¹× Á¤Ã¥ ÃÖÀûÈ °úÁ¤À» ¹Ýº¹ÇÏ´Â Closed-Loop ±¸Á¶·Î ÀÌ·ç¾îÁø ¾Ë°í¸®ÁòÀÌ´Ù. ÀÌ·¯ÇÑ °ÈÇнÀÀÇ ÁÖ¿ä ÀåÁ¡Àº ¾×¼ÇÀÇ Ç°ÁúÀ» Æò°¡ÇÏ°í Á¤Ã¥À» Áö¼ÓÀûÀ¸·Î ÃÖÀûÈ ÇÏ´Â °ÍÀÌ´Ù. µû¶ó¼, °ÈÇнÀÀº Áö´ÉÇü ½Ã½ºÅÛ, ÀÚÀ²Á¦¾î ½Ã½ºÅÛ °³¹ß¿¡ È¿°úÀûÀ¸·Î È°¿ëµÉ ¼ö ÀÖ´Ù. ±âÁ¸ÀÇ °ÈÇнÀÀº, ´ÜÀÏ Á¤Ã¥, ´ÜÀÏ º¸»óÇÔ¼ö ¹× ºñ±³Àû ´Ü¼øÇÑ Á¤Ã¥ ¾÷µ¥ÀÌÆ® ±â¹ýÀ» Á¦ÇÑÀûÀÎ ¹®Á¦¿¡ ´ëÇØ Á¦½ÃÇÏ°í Àû¿ëÇÏ¿´´Ù. º» ³í¹®¿¡¼´Â ±¸¼º¿ä¼ÒÀÇ º¹¼ö¼ºÀ» Áö¿øÇÏ´Â È®ÀåµÈ °ÈÇнÀ ¸ðµ¨À» Á¦¾ÈÇÑ´Ù. Á¦¾ÈµÇ´Â È®Á¤ °ÈÇнÀÀÇ ÁÖ¿ä ±¸¼º ¿ä¼ÒµéÀ» Á¤ÀÇÇÏ°í, ±×µéÀÇ ÄÄÇ»Æà ¸ðµ¨À» Æ÷ÇÔÇÏ´Â Á¤Çü ¸ðµ¨À» Á¦½ÃÇÑ´Ù. ¶ÇÇÑ, ÀÌ Á¤Çü¸ðµ¨À» ±â¹ÝÀ¸·Î ½Ã½ºÅÛ °³¹ßÀ» À§ÇÑ ¼³°è ±â¹ýÀ» Á¦½ÃÇÑ´Ù. Á¦¾ÈÇÑ ¸ðµ¨À» ±â¹ÝÀ¸·Î ÀÚÀ² ÃÖÀûÈ ÀÚµ¿Â÷ ³»ºñ°ÔÀÌÅÍ ½Ã½ºÅÛ¿¡ Àû¿ë ¹× ½ÇÇèÀ» ÁøÇàÇÑ´Ù. Á¦½ÃµÈ Á¤Çü ¸ðµ¨°ú ¼³°è ±â¹ýÀ» Àû¿ëÇÑ »ç·Ê¿¬±¸·Î, º¹¼öÀÇ ÀÚµ¿Â÷µéÀÌ ÃÖÀû ¸ñÀûÁö¿¡ ´Ü ½Ã°£¿¡ µµÂøÇÒ ¼ö ÀÖ´Â ÁøÈµÈ ³»ºñ°ÔÀÌÅÍ ½Ã½ºÅÛ ¼³°è ¹× ±¸ÇöÀ» ÁøÇàÇÑ´Ù. |
¿µ¹®³»¿ë (English Abstract) |
Reinforcement Learning (RL) is a machine learning algorithm that repeat the closed-loop process that agents perform actions specified by the policy, the action is evaluated with a reward function, and the policy gets updated accordingly. The key benefit of RL is the ability to optimze the policy with action evaluation. Hence, it can effectively be applied to developing advanced intelligent systems and autonomous systems. Conventional RL incoporates a single policy, a reward function, and relatively simple policy update, and hence its utilization was limited. In this paper, we propose an extended RL model that considers multiple instances of RL elements. We define a formal model of the key elements and their computing model of the extended RL. Then, we propose design methods for applying to system development. As a case stud of applying the proposed formal model and the design methods, we present the design and implementation of an advanced car navigator system that guides multiple cars to reaching their destinations efficiently. |
Å°¿öµå(Keyword) |
³»ºÎÀüÆÄ°æ·Î ŽÁö
ÆäÀÌÁö·©Å© ¾Ë°í¸®Áò
¼³¸í°¡´ÉÇÑ ÀΰøÁö´É
¿ø°Ý µ¥½ºÆ®Åé ÇÁ·ÎÅäÄÝ
Ư¡ ÃßÃâ
Lateral Movement
Pagerank Algorithm
Explainable AI
Remote Desktop Protocol
Feature Extraction
°ÈÇнÀ
È®ÀåµÈ °ÈÇнÀ ¸ðµ¨
Á¤Çü ¸ðµ¨
¼³°è ±â¹ý
ÁøÈµÈ ³×ºñ°ÔÀÌÅÍ ½Ã½ºÅÛ
Reinforcement Learning (RL)
Advanced RL
Formal Model
Design Methods
Advanced Navigator System
|
ÆÄÀÏ÷ºÎ |
PDF ´Ù¿î·Îµå
|