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

»çÀÌÆ®¸Ê

Loading..

Please wait....

±¹³» ³í¹®Áö

Ȩ Ȩ > ¿¬±¸¹®Çå > ±¹³» ³í¹®Áö > Çѱ¹Á¤º¸°úÇÐȸ ³í¹®Áö > Á¤º¸°úÇÐȸ³í¹®Áö (Journal of KIISE)

Á¤º¸°úÇÐȸ³í¹®Áö (Journal of KIISE)

Current Result Document :

ÇѱÛÁ¦¸ñ(Korean Title) Â÷º°Àû ¼Õ½ÇÀ» ÀÌ¿ëÇÑ ¸ðµ¨±â¹Ý °­È­ÇнÀ
¿µ¹®Á¦¸ñ(English Title) Model-Based Reinforcement Learning with Discriminative Loss
ÀúÀÚ(Author) Áø ±¤   ³ë¿äȯ   À̵µÈÆ   Guang Jin   Yohwan Noh   DoHoon Lee  
¿ø¹®¼ö·Ïó(Citation) VOL 47 NO. 06 PP. 0547 ~ 0552 (2020. 06)
Çѱ۳»¿ë
(Korean Abstract)
°­È­ÇнÀÀº ¿©·¯ °¡Áö ¾î·Á¿î ¹®Á¦µéÀ» ÇØ°áÇÏ´Â µ¥ ÁÁÀº °á°ú¸¦ º¸¿©ÁÖ°í ÀÖ´Ù. ±×·¯³ª À̸¦ ½ÇÁ¦ ¹®Á¦¿¡ Àû¿ëÇϱ⿡´Â »ùÇà ȿÀ²¼ºÀÌ Å« ¹®Á¦ÀÌ´Ù. ÀÌ ³í¹®¿¡¼­´Â Â÷º° ¼Õ½ÇÇÔ¼ö¸¦ ÀÌ¿ëÇÑ ¸ðµ¨±â¹Ý °­È­ÇнÀ ÇÁ·¹ÀÓ¿öÅ©¸¦ Á¦¾ÈÇÑ´Ù. ÀÌ ¹æ¹ýÀº ¸ðµ¨ÀÌ ¼­·Î ´Ù¸¥ µ¿ÀÛÀ» ±¸º°ÇÒ ¼ö ÀÖµµ·Ï ÈÆ·ÃÇÑ´Ù. ÀÌ ÇÁ·¹ÀÓ¿öÅ©·Î »çÀü ÇнÀµÈ ÀÎÄÚ´õ°¡ ÃßÃâÇÑ Æ¯Â¡Àº Á¤Ã¥ ±×¶óµð¾ðÆ® ¹æ¹ýÀÌ ÃßÃâÇÑ Æ¯Â¡°ú ÀÏÄ¡ÇÑ´Ù´Â °ÍÀ» ¹ß°ßÇß´Ù. Á¦¾ÈÇÑ ¹æ¹ýÀº ¾ÆŸ¸®(Atari) °ÔÀÓ È¯°æ¿¡¼­ ±âÁ¸ÀÇ ¸ðµ¨±â¹Ý °­È­ÇнÀ ¹æ¹ýº¸´Ù ³ôÀº »ùÇà ȿÀ²¼ºÀ» º¸¿´À¸¸ç ƯÈ÷ ÇнÀÀÇ Ãʱ⠴ܰ迡¼­´Â ±âÁؼ±º¸´Ùµµ ³ôÀº È¿À²¼ºÀ» º¸¿´´Ù
¿µ¹®³»¿ë
(English Abstract)
Reinforcement learning is a framework for training the agent to make a good sequence of decisions through interacting with a complex environment. Although reinforcement learning has shown promising results in many tasks, sample efficiency still remains a major challenge for its real world application. We propose a novel model-based reinforcement learning framework that incorporates the discriminative loss function, in which models are trained to discriminate one action from another. The encoder pre-trained in this framework shows the feature alignment property, which aligns with the policy gradient method. The proposed method showed better sample efficiency than conventional model-based reinforcement learning approaches in the Atari game environment. In the early stage of the training, the proposed method surpassed the baseline by a large margin
Å°¿öµå(Keyword) °­È­ÇнÀ   ¸ðµ¨±â¹Ý   Â÷º° ¼Õ½Ç ÇÔ¼ö   µ¿ÀÛ ±¸º°   reinforcement learning   model-based RL   discriminative loss function   action discriminatio  
ÆÄÀÏ÷ºÎ PDF ´Ù¿î·Îµå