Loading...

Çмú´ëȸ

Korea Robotics Society

Á¦2ȸ ÀΰøÁö´É ¹× ·Îº¸Æ½½º ¿©¸§Çб³
AI & Robotics Summer School 2017

  >   Çмú´ëȸ   >   ÀΰøÁö´É ¹× ·Îº¸Æ½½º ¿©¸§/°Ü¿ïÇб³

ÀΰøÁö´É ¹× ·Îº¸Æ½½º ¿©¸§/°Ü¿ïÇб³

ÇÁ·Î±×·¥

Çà»ç°³¿ä

Çà»ç¸í Á¦2ȸ ÀΰøÁö´É ¹× ·Îº¸Æ½½º ¿©¸§Çб³ (AI & Robotics Summer School 2017)
ÀϽà 2017³â 8¿ù 16ÀÏ(¼ö) ~ 18ÀÏ(±Ý), 3ÀÏ°£
Àå¼Ò ´ëÀü KAIST ´Ù¸ñÀûȦ (´ëÀü±¤¿ª½Ã À¯¼º±¸ ´ëÇзΠ291, N1 ±èº´È£¤ý±è»ï¿­ ITÀ¶ÇÕ ºôµù)
ÁÖÃÖ/ÁÖ°ü Çѱ¹·Îº¿ÇÐȸ
ÃëÁö ±¹³» ´ëÇпø»ýµéÀ» Æ÷ÇÔÇÑ robotics community¿¡ AI & robotics ºÐ¾ßÀÇ ¸í°­ÀÇ¿Í ½Ç½ÀÀ» Æ÷ÇÔÇÑ Æò¼Ò Á¢Çϱ⠾î·Á¿ü´ø ±³À° ¼­ºñ½º¸¦ Á¦°øÇÑ´Ù.

¡Ø 2016³âµµ¿ÍÀÇ Â÷º°¼º ¡Ø
´ëÇпø»ýµéÀ» À§ÇÑ ±âÃÊÀûÀÎ °­ÀÇ¿¡ ÁßÁ¡À» µÎ¸ç, ÃֽŠÅäÇÈÀ» ¼Ò°³ÇÏ´Â ÇÁ·Î±×·¥À¸·Î ±¸¼º ÇÔ.
½Ç½ÀÀ» ÅëÇØ Á÷Á¢ üÇèÇÒ ¼ö Àִ ȯ°æÀ» Á¦°ø ¿¹Á¤

ÇÁ·Î±×·¥(¾È)

°­ÀÇ ½Ã°£Ç¥(¾È)

Àå¼Ò: KAIST ´Ù¸ñÀûȦ (N1 ±èº´È£¤ý±è»ï¿­ ITÀ¶ÇÕ ºôµù)
  8/16(¼ö) 8/17(¸ñ) 8/18(±Ý)
09:00-10:00 µî·Ï µö·¯´×
(KAIST ±èÁظ𠱳¼ö)
µö·¯´× ½Ç½À
(KAIST ±èÁظ𠱳¼ö)
10:00-12:30 ¸Ó½Å·¯´×ÀÇ ¼Ò°³
(KIAST À¯Ã¢µ¿ ±³¼ö)
10:00~13:00
12:30-14:00 Áß½Ä
±³¼öȸ°ü [°Ç¹°¹øÈ£ N6 2Ãþ]
14:00-17:30 Natural Language Understanding
(ÀüºÏ´ë ³ª½ÂÈÆ ±³¼ö)
µö·¯´× ½Ç½À
(KAIST ±èÁظ𠱳¼ö)
ÄÄÇ»ÅͺñÀü: À̷аú ÀÀ¿ë
(KAIST ±ÇÀÎ¼Ò ±³¼ö)
¡Ø °­ÀÇ ½Ã°£ ¹× Á¦¸ñÀº º¯µ¿µÉ ¼ö ÀÖÀ½

°­ÀÇ°³¿ä

°­»ç À¯Ã¢µ¿ ±³¼ö (KAIST)
Á¦¸ñ ÀΰøÁö´É °³·Ð, ¸Ó½Å·¯´× °³·Ð
ÃÊ·Ï º» °­¿¬¿¡¼­´Â Machine Learning (±â°èÇнÀ)ÀÇ ±âº» °³³ä¿¡ ´ëÇؼ­ ¾Ë¾Æº»´Ù. ±â°è ÇнÀÀÇ ±âº» ¿øÄ¢°ú ü°è ±×¸®°í ´Ù¾çÇÑ ºÐ¾ß¸¦ ¼Ò°³ÇÏ°í Bayesian principle, EM algorithms, ±×¸®°í Support Vector Machine¿¡ ´ëÇؼ­ ¹è¿î´Ù. °­ÀÇ ³»¿ëÀº ´ÙÀ½°ú °°´Ù.
  1. Introduction to Machine Learning
  2. Probability Distributions
  3. Bayesian Approach
  4. Expectation-Maximization Algorithm(EM Algorithm)
  5. Support Vector Machine(SVM)
¾à·Â 1986 B.S. degree in Engineering and Applied Science/California Institute of Technology
1988 M.S. degree in Electrical Engineering/Cornell University
1996 Ph. D. degree in Electrical Engineering/Massachusetts Institute of Technology

1999 – ÇöÀç KAIST Àü±â¹×ÀüÀÚ°øÇкΠ±³¼ö
1997 – 1999 Çѱ¹Åë½Å(KT) ¿¬±¸°³¹ßº»ºÎ ¼±ÀÓ¿¬±¸¿ø
 
°­»ç ³ª½ÂÈÆ ±³¼ö (ÀüºÏ´ëÇб³ ÄÄÇ»ÅÍ°øÇкÎ)
Á¦¸ñ Natural Language Understanding
ÃÊ·Ï º» °­ÀÇ¿¡¼­´Â µö ·¯´×À» ÀÌ¿ëÇÑ ÀÚ¿¬¾ð¾îÀÌÇØÀÇ Àü¹Ý¿¡ ´ëÇؼ­ ´Ù·ç¸ç, °­ÀÇ ³»¿ëÀº ´ÙÀ½°ú °°´Ù.
  1. Introduction to natural language understanding
  2. Word embedding and neural language models
  3. Recurrent/Recursive neural networks
  4. Neural encoder-decoder for machine translation
  5. Transition-based neural networks
  6. Memory-augmented neural networks
¾à·Â 2008: POSTECH ÄÄÇ»ÅÍ°øÇаú ¹Ú»ç
2008-2012: National University of Singapore, Research fellow
2012-2014: Çѱ¹ÀüÀÚÅë½Å¿¬±¸¿ø, ¼±ÀÓ¿¬±¸¿ø
2015-: ÀüºÏ´ëÇб³ ÄÄÇ»ÅÍ°øÇкΠÁ¶±³¼ö
 
°­»ç ±èÁظ𠱳¼ö (Ä«À̽ºÆ® Àü±â¹×ÀüÀÚ°øÇаú)
Á¦¸ñ µö·¯´×
ÃÊ·Ï µö·¯´×Àº ÃÖ±Ù À½¼º ÀνÄ, ¿µ»ó ÀÎ½Ä ºÐ¾ßÀÇ °¢Á¾ ¼¼°è ±â·ÏÀ» »õ·Î ¼ö¸³Çϸ鼭 °­·ÂÇÑ ±â°èÇнÀ ¹æ¹ýÀ¸·Î °¢±¤¹Þ°í ÀÖ´Ù. ƯÈ÷ ±âÁ¸¿¡ »ç¶÷ÀÌ ¼öµ¿À¸·Î °¢Á¾ feature¸¦ designÇÑ ÈÄ ±â°èÇнÀ ¹æ¹ý°ú °áÇÕÇÏ¿© ºÐ·ù, ÀÎ½Ä ¹®Á¦¸¦ ÇØ°áÇÏ´ø Æз¯´ÙÀÓÀ» Å»ÇÇÇÏ¿© data·ÎºÎÅÍ ÀÚµ¿ÀûÀ¸·Î °èÃþÀûÀÎ featureµéÀ» ÇнÀÇÏ°í ºÐ·ù, ÀνıîÁö ÅëÇÕÇÏ¿© ¼öÇàÇÒ ¼ö ÀÖ´Ù´Â Á¡¿¡¼­ µö·¯´×Àº ±â°èÇнÀÀÇ »õ·Î¿î Æз¯´ÙÀÓÀ» Á¦½ÃÇÏ¿´´Ù°í ÇÒ ¼ö ÀÖ´Ù. º» °­ÀÇ¿¡¼­´Â µö·¯´×ÀÌ ¹«¾ùÀ̸ç, ¿Ö ±× µ¿¾È µö·¯´×ÀÌ ¾î·Á¿ü´ÂÁö »ìÆ캸°í deep learningÀ» °¡´ÉÇÏ°Ô ÇÑ ÃÖ±Ù ¿¬±¸ ¼º°úµéÀ» ¼Ò°³ÇÑ´Ù. ÀÌ¿Í °ü·ÃÇÏ¿© restricted Boltzmann machine (RBM), deep belief network (DBN), deep neural network (DNN), convolutional neural network (CNN) µîÀÇ ±â¹ýÀ» ¼Ò°³ÇÏ°í À̵éÀÌ ¾ó±¼ ÀνÄ, ¹°Ã¼ ÀνÄÀ» Æ÷ÇÔÇÑ robot vision ¹®Á¦µé¿¡ ¾î¶»°Ô È°¿ëµÇ´ÂÁö ¼Ò°³ÇÑ´Ù. ¶ÇÇÑ ½Ç½À ½Ã°£¿¡´Â Torch, Pytorch, Tensorflow, Caffe¿Í °°Àº µö·¯´×À» À§ÇÑ ¶óÀ̺귯¸®ÀÇ È°¿ë ¹æ¹ýÀ» »ìÆ캸°í Á÷Á¢ »ç¿ëÇØ º¸´Â ½Ç½À ±âȸ¸¦ Á¦°øÇÑ´Ù.

< ½Ç½À ³»¿ë ¹× ´ã´ç Á¶±³ >
- Caffe ¼³Ä¡ ¹× »ç¿ë¹æ¹ý, Caffe ±¸Á¶ ÆÄ¾Ç ¹× CNNÀ» ±¸¼ºÇÏ°í ÇнÀ½ÃÅ°´Â ¹æ¹ý ½Ç½À (ÁÖµ¿±Ô)
- TensorflowÀÇ ±âº»ÀûÀÎ ½ÇÇà ¹× Examples, Tensorboard¸¦ ÀÌ¿ëÇÑ °á°ú visualization (ÀÌÀ翵)
- TorchÀÇ lua¸¦ ÀÌ¿ëÇÑ ±âÃÊ ÇÁ·Î±×·¡¹Ö ¼³¸í, CNN ±¸Çö ¹× ÇнÀ ¹æ¹ý ½Ç½À (±è¿µµ¿)
- PytorchÀÇ ±âº»ÀûÀÎ »ç¿ë¹æ¹ý ¹× ±¸Á¶ ÆľÇ, °£´ÜÇÑ ¿¹Á¦ ½Ç½À (ÀÌÁ¾È£)
¾à·Â 1995-1998: ¼­¿ï´ëÇб³ Àü±â°øÇкΠÇлç
1998-2005: MIT, EECS¼®»ç, ¹Ú»ç
2005-2009:»ï¼ºÁ¾ÇÕ±â¼ú¿¬±¸¿ø Àü¹®¿¬±¸¿ø
2009-ÇöÀç: KAIST Àü±â¹×ÀüÀÚ°øÇаú Á¶±³¼ö, ºÎ±³¼ö
 
°­»ç ±ÇÀÎ¼Ò ±³¼ö (Ä«À̽ºÆ® Àü±â¹×ÀüÀÚ°øÇкÎ)
Á¦¸ñ ÄÄÇ»ÅͺñÀü: À̷аú ÀÀ¿ë
ÃÊ·Ï ÄÄÇ»ÅͺñÀüÀº Àΰ£½Ã°¢½Ã½ºÅÛÀÇ ÀÛµ¿¿ø¸®¸¦ ¼öÇÐÀû¸ðµ¨·Î ºÐ¼®ÇÏ°í ÀÌÇØÇÏ·Á´Â ½ÃµµµéÀ» Áß½ÉÀ¸·Î ¹ßÀü ÇÏ¿©¿Ô´Ù. ÃÖ±Ù µé¾î¼­´Â ÄÄÇ»Åͱ׷¡ÇȽº, ¸Ó½Å·¯´× µî »õ·Î¿î ºÐ¾ß¿ÍÀÇ À¶ÇÕÀ¸·Î ±Þ¼ÓÀûÀÎ ¹ßÀüÀ» ÀÌ·ç¾ú´Ù. º» °­¿¬¿¡¼­´Â ÄÄÇ»ÅͺñÀü¿¡ ´ëÇÑ ±âÃÊ¿¡¼­ºÎÅÍ ´Ù¾çÇÑ ÀÀ¿ë±îÁö¸¦ ¼Ò°³ÇÑ´Ù. ±¸Ã¼ÀûÀ¸·Î´Â (1) Ä«¸Þ¶ó ¸ðµ¨¿¡ ´ëÇÑ ¼Ò°³ ¹× ¼öÇÐÀû ¸ðµ¨, (2) ¿©·¯ ´ëÀÇ Ä«¸Þ¶ó¸¦ ÀÌ¿ëÇÑ 3Â÷¿øº¹¿ø¿¡ ´ëÇÑ ±âº» ÀÌ·Ð, (3) ¹à±â°ªÀ» ÀÌ¿ëÇÑ 3Â÷¿ø Á¤º¸ ÃßÃâ¿¡ ´ëÇÑ ±âº» ¹æ¹ý·Ð°ú ±× ÀÌ·Ð, (4) Æ®·¡Å·, ¹°Ã¼ºÐÇÒ µî°ú °°Àº ±âº» ¾Ë°í¸®µë¿¡ ´ëÇÑ ¼Ò°³, (5) Ư¡ ÃßÃâ°ú ¸ÅĪ, (6) ¹°Ã¼ÀÎ½Ä ¹æ¹ý·Ð µîÀ» ¼Ò°³ÇÑ´Ù. ƯÈ÷, °¢ ÁÖÁ¦ º°·Î ±âÁ¸ ¹æ¹ý·ÐµéÀÌ °®°í ÀÖ´Â ÇѰ踦 ¼Ò°³ÇÏ°í, ÃÖ±Ù µö·¯´× ±â¹ÝÀÇ ¹æ¹ý·ÐµéÀÌ ÀÌ·¯ÇÑ ¹®Á¦¸¦ ÇØ°áÇÏ´Â »õ·Î¿î Á¢±Ù¹ýµéÀ» ºñ±³ÇÏ¿© ¼³¸íÇÑ´Ù.
¾à·Â ¼­¿ï´ëÇб³ ±â°è¼³°èÇаú, °øÇлç, 1981
¼­¿ï´ëÇб³ ±â°è¼³°èÇаú, °ø¼®»ç, 1983
Carnegie Mellon University, Ph.D. in Robotics, 1990

Çѱ¹±â°è¿¬±¸¿ø, ¿¬±¸¿ø, 1983~1984
Toshiba R&D Center, Researcher, 1991~1992
Cambridge University, Visiting Professor, 1998~1999
Ä«À̽ºÆ® Àü±â¹×ÀüÀÚ°øÇкÎ, ±³¼ö, 1992~ÇöÀç
Ä«À̽ºÆ® Àü±â¹×ÀüÀÚ°øÇкÎ, ÇÑÀü¼®Á±³¼ö, 2015~ÇöÀç