1. ÀÌ Ã¥Àº Çпø ¼ö°À» ÅëÇØ ½ÀµæÇÑ ³»¿ë°ú ¸àÅ丵À» ÁøÇàÇÏ¸é¼ ½º½º·Î ÇнÀÇÑ ³»¿ëÀ» ¹ÙÅÁÀ¸·Î ´ä¾È ÇüÅ·ΠÀÛ¼ºÇÏ¿´°í, ITºÐ¾ß ±â¼ú»çÀÎ Á¤º¸°ü¸®±â¼ú»ç¿Í ÄÄÇ»ÅͽýºÅÛÀÀ¿ë±â¼ú»ç ÀÚ°ÝÀ» ÃëµæÇϱâ À§ÇØ ÇнÀÇÏ°í Àְųª ÇнÀÇÏ°íÀÚ ÇÏ´Â ºÐµéÀ» À§ÇØ ¸¸µé¾ú½À´Ï´Ù.2. º» ±³Àç´Â ¹ßÀü µ¿Çâ, ¹è°æ ±×¸®°í À¯»ç ±â¼ú°úÀÇ ºñ±³, ´Ù¾çÇÑ µµ½ÄÈ µî 25³â°£ÀÇ ½Ç¹« °³¹ßÀÚ °æÇèÀ» Åä´ë·Î ÀÛ¼ºÇÑ ³»¿ëÀ¸·Î dzºÎÇÑ °æÇèÀûÀÎ ¿ä¼Ò°¡ ³»ÀçµÇ¾î ÀÖ´Â ÀåÁ¡ÀÌ ÀÖ½À´Ï´Ù.
¼º±Õ°ü´ëÇб³ Á¤º¸Åë½Å´ëÇпø Á¤º¸º¸È£Çаú Á¹¾÷(°øÇм®»ç)
»ï¼ºÀüÀÚ ¼±ÀÓ/Ã¥ÀÓ ¿¬±¸¿ø
µµ½Ã¹Ù »ï¼º ½ºÅ丮Áö Å×Å©³î·¯Áö ÄÚ¸®¾Æ(ÁÖ) ¼ö¼®¿¬±¸¿ø
ÄÄÇ»ÅͽýºÅÛÀÀ¿ë ±â¼ú»ç, Á¤º¸½Ã½ºÅÛ ¼ö¼®°¨¸®¿ø, Á¤º¸Åë½Å Ư±Þ±â¼úÀÚ
°úÇбâ¼úÁ¤º¸Åë½ÅºÎ IT ¸àÅä
µ¥ÀÌÅÍ°ü¸®ÀÎÁõ½É»ç¿ø(DQC-M)
ùÛ(ÇÑ)¡¤ìí(ÀÏ)±â¼ú»ç ±³·ùȸ À§¿ø
http://cafe.naver.com/96starpe ¿î¿µÀÚ
PART 1. ÀΰøÁö´É(ìÑÍïò±Òö, Artificial Intelligence)ÀÇ °³¿ä1. ÀΰøÁö´É(Artificial Intelligence)ÀÇ ¿ª»ç2. ÀΰøÁö´É3. ¾à ÀΰøÁö´É(Weak AI), ° ÀΰøÁö´É(Strong AI), ÃÊ ÀΰøÁö´É(Super AI)4. ÀΰøÁö´É(AI)ÀÇ Æ¯ÀÌÁ¡(Singularity)5. ¾Æ½Ç·Î¸¶(ASILOMA) AI(ÀΰøÁö´É) ¿øÄ¢6. ±ÔÄ¢±â¹Ý¸ðµ¨7. Ãßõ¿£Áø(Recommendation Engine)8. Àü¹®°¡½Ã½ºÅÛ(Expert System)9. Á¤±ÔÇ¥Çö½Ä°ú À¯ÇÑ ¿ÀÅ丶Ÿ10. À¯ÇÑ ¿ÀŸ¸¶Å¸(Finite Automata)11. Æ©¸µÅ×½ºÆ®(Turing Test)12. ¿¡ÀÌÀüÆ®(Agent) - 1±³½ÃÇü ´ä¾È13. ¿¡ÀÌÀüÆ®(Agent) - 2±³½ÃÇü ´ä¾È14. ų ½ºÀ§Ä¡(Kill Switch)15. Æ®·Ñ¸® µô·¹¸¶(Trolley Dilemma)16. ÀΰøÁö´É(AI) À±¸®ÀÇ °³³ä, ÁÖ¿ä »ç·Ê, °í·Á»çÇ× ¹× ÃßÁø ¹æÇâ17. ÀÌ¿ëÀÚ Áß½ÉÀÇ Áö´ÉÁ¤º¸»çȸ¸¦ À§ÇÑ ¿øÄ¢PART 2. ÀΰøÁö´É ¾Ë°í¸®Áò(Algorithm)18. À¯ÀüÀÚ ¾Ë°í¸®Áò(Genetic Algorithm)19. ±×¸®µð ¾Ë°í¸®Áò(Greedy Algorithm)20. »ó°üºÐ¼®(Correlation Analysis)21. ȸ±ÍºÐ¼®(Regression Analysis)22. ·ÎÁö½ºÆ½ ȸ±ÍºÐ¼®(Logistic Regression Analysis) 23. ±ºÁýºÐ¼®(Cluster Analysis) - 1±³½ÃÇü ´ä¾È24. ±ºÁýºÐ¼®(Cluster Analysis) - 2±³½ÃÇü ´ä¾È25. °èÃþÀû ±ºÁýºÐ¼®(Hierarchical Clustering)26. ÀÚÄ«µå(Jaccard)°è¼ö27. ÇعְŸ®(Hamming Distance)28. À¯Å¬¸®µð¾È °Å¸®(Euclidean Distance)29. À¯Å¬¸®µð¾È °Å¸®(Euclidean Distance)¸¦ °è»êÇϽÿÀ.30. ¸¶ÇÒ¶ó³ëºñ½º °Å¸®(Mahalanobis Distance)¸¦ °è»êÇϽÿÀ.31. Apriori(¿¬°ü±ÔÄ¢) ¾Ë°í¸®Áò32. ÁöÁöµµ(Support), ½Å·Úµµ(Confidence), Çâ»óµµ(Lift)33. »ç·Ê1(TV ±¸ÀԽà DVD ±¸ÀÔ), »ç·Ê2(¿ìÀ¯ ±¸ÀԽà ÁÖ½º ±¸ÀÔ)¿¡ ´ëÇØ ¿¬°ü±ÔÄ¢(ÁöÁöµµ, ½Å·Úµµ, Çâ»óµµ)À» Á¦½ÃÇϽÿÀ.34. ¾Ó»óºíÇнÀ(Ensemble Learning)35. ¸Ó½Å·¯´×(Machine Learning)¿¡ È°¿ë, ¾Ó»óºí(Ensemble) ±â¹ý36. Bagging°ú Boosting ºñ±³37. ·£´ý Æ÷·¹½ºÆ®(Random Forest)38. ÀÇ»ç°áÁ¤Æ®¸®(Decision Tree)39. K-NN(K-Nearest Neighbor)40. ½Ã°è¿ ºÐ¼®41. ½Ã°è¿ ºÐ¼®(ARIMA)42. SVM(Support Vector Machine)- 1±³½ÃÇü ´ä¾È43. SVM(Support Vector Machine)- 2±³½ÃÇü ´ä¾È44. º£ÀÌÁî(Bayes)Á¤¸®45. Å©±â¿Í ¸ð¾çÀÌ °°Àº °øÀÌ »óÀÚ A¿¡´Â °ËÀº °ø 2°³¿Í Èò°ø 2°³, »óÀÚ B¿¡´Â °ËÀº°ø 1°³¿Í Èò°ø 2°³°¡ µé¾î ÀÖ´Ù. µÎ »óÀÚ A, B Áß ÀÓÀÇ·Î ¼±ÅÃÇÑ ÇϳªÀÇ »óÀÚ¿¡¼ °øÀ» 1°³ ²¨³Â´õ´Ï °ËÀº°øÀÌ ³ª¿ÔÀ» ¶§, ±× »óÀÚ¿¡ ³²Àº °øÀÌ ¸ðµÎ Èò°øÀÏ È®·üÀº? (º£ÀÌÁî(Bayes)Á¤¸®¸¦ È°¿ëÇϽÿÀ)46. K-Means47. DBSCAN(Density Based Spatial Clustering with Application Notes)48. Â÷¿øÃà¼Ò(Dimensionality Reduction)49. Ư¡ÃßÃâ(Feature Extraction)50. ÁÖ¼ººÐ ºÐ¼®, PCA(Principal Component Analysis)51. µ¶¸³¼ººÐºÐ¼®, ICA(Independent Component Analysis)52. ¸¶¸£ÄÚÇÁ °áÁ¤ ÇÁ·Î¼¼½º(Markov Decision Process, MDP)53. Àº´Ð ¸¶¸£ÄÚÇÁ ¸ðµ¨(HMM-Hidden Markov Model)54. ¸óÅ×Ä«¸¦·Î Æ®¸® Ž»ö(MCTS)55. Q-Learning56. Tokenization(Åä±ÙÈ), N-gram57. Word2Vec58. Word2VecÇнÀ¸ðµ¨, CBOW(Continuous Bag Of Words), Skip-gramPART 3. ½ÉÃþ ½Å°æ¸Á »ó¼¼59. ÀϹÝÀûÀÎ ÇÁ·Î±×·¥ ¹æ½Ä°ú Machine Learning(±â°èÇнÀ) ÇÁ·Î±×·¡¹Ö ¹æ½Ä60. AI(Artificial Intelligence), ML(Machine Learning), DL(Deep Learning)61. ±â°èÇнÀ(Machine Learning)62. ÁöµµÇнÀ(Supervised Learning)63. ºñÁöµµ(ºñ°¨µ¶)(Unsupervised Learning)ÇнÀ64. °ÈÇнÀ(Reinforcement Learning)65. µö·¯´×(Deep Learning)66. MCP(McCulloch-Pitts)´º·±(Neuron)67. Çñ ±ÔÄ¢(Hebb Rule)68. ÆÛ¼ÁÆ®·Ð(Perceptron)69. ¾Æ´Þ¶óÀÎ(Adaline- Adaptive Linear Neutron)70. È°¼ºÈ ÇÔ¼ö(Activation Function) - 171. È°¼ºÈ ÇÔ¼ö(Activation Function) - 272. ½Å°æ¸Á ÇнÀ - FFNN(Feed Forward Neural Network)73. µö·¯´×(Deep Learning)ÀÇ ÆĶó¹ÌÅÍ(Parameter)¿Í ÇÏÀÌÅÍÆĶó¹ÌÅÍ (Hyperparameter)¸¦ ºñ±³ÇÏ°í ÇÏÀÌÆÛÆĶó¹ÌÅÍÀÇ Æ©´×¹æ¹ýÀ» ¼³¸íÇϽÿÀ74. ¿ªÀüÆĹý(Back-Propagation)75. ±â¿ï±â ¼Ò½Ç ¹®Á¦(Vanishing Gradient Problem)76. °æ»çÇÏ°¹ý(Gradient Descent)77. °úÀûÇÕ(Overfitting)°ú ºÎÀûÇÕ(Underfitting), ÀûÇÕ(Bestfitting)78. °úÀûÇÕ(Overfitting)°ú ºÎÀûÇÕ(Underfitting) ÇØ°á¹æ¾È79. Dropout80. ANN(Artificial Neural Network)81. DNN(Deep Neural Network)82. CNN(Convolution Neural Network)83. RNN(Recurrent Neural Network)84. LSTM(Long Short-Term Memory)85. GRU(Gated Recurrent Unit)86. RBM(Restricted Boltzmann Machine)87. DBN(Deep Belief Network)88. DQN(Deep Q-Network)89. GAN(Generative Adversarial Networks) [ GANÀÇ ÀÌÇØ ]90. DL4J(Deep Learning 4J)91. È¥µ¿Çà·Ä(Confusion Matrix)92. Machine Learning(±â°èÇнÀ)ÀÇ Æò°¡¹æ¹ý-Accuracy(Á¤È®µµ), Recall(ÀçÇöÀ²), Precision(Á¤¹Ðµµ)93. F1 ScorePART 4. ÀΰøÁö´É È°¿ë94. À½¼ºÀνıâ¼ú, ASR(Automatic Speech Recognition), NLU(Natural Language Understanding) TTS(Text to Speech)95. À½¼ºÀνÄ(Voice Recognition)96. 꺿(ChatBot)97. °¡»ó°³Àκñ¼(Virtual Personal Assistant)98. ÆÐÅÏÀνÄ(Pattern Recognition)99. ¸Ó½Å·¯´× ÆÄÀÌÇÁ¶óÀÎ(Machine Learning Pipeline)100. ÀÚ¿¬¾î ó¸®101. ¿¢¼Òºê·¹ÀÎ(Exobrain)102. ¿¢¼Òºê·¹ÀÎ(Exobrain)°ú Deepview ±â¼ú¿ä¼Ò103. µöºä(Deepview)104. SNA(Social Network Analysis)105. ÅÙ¼Ç÷Î(Tensorflow)106. ÆÄÀ̼Ç(Python)ÀÇ Æ¯Â¡ ¹× ÀÚ·áÇü(Data Type)107. ÆÐ¼Ç ÀÇ·ù¿ë À̹ÌÁö¸¦ ºÐ·ùÇÏ´Â ´ÙÃþ ½Å°æ¸ÁÀ» µé·Á°í ÇÑ´Ù. ÀÇ·ù¿ë À̹ÌÁö´Â ¹ÙÁö, Ä¡¸¶, ¼ÅÃ÷ µî 10°¡Áö À¯ÇüÀÇ Èæ¹é À̹ÌÁö(32*32 pixels)·Î ±¸¼ºµÇ¾î ÀÖ°í, ÇнÀ¿¡ ÅõÀÔÇÒ À̹ÌÁö µ¥ÀÌÅÍ´Â °ËÁõ ¹× Å×½ºÆ®¿ë µ¥ÀÌÅ͸¦ Á¦¿ÜÇÏ°í ÃÑ 48,000ÀåÀÌ´Ù. ÀÔ·ÂÃþ, Àº´ÐÃþ, Ãâ·ÂÃþÀÇ ¿ÏÀü¿¬°á(fully connected) 3°èÃþÀ¸·Î ±¸¼ºµÇ¾î ÀÖ°í Àº´ÐÃþÀÇ ´º·±°³¼ö´Â 100°³ÀÏ ¶§, ´ÙÀ½¿¡ ´ëÇÏ¿© ¼³¸íÇϽÿÀ °¡. ½Å°æ¸Á ±¸¼ºµµ ³ª. ÀÔ·ÂÃþÀÇ ÀԷ°³¼ö, Ãâ·ÂÃþÀÇ ´º·± °³¼ö, ÇнÀÇÒ °¡ÁßÄ¡¿Í ÀýÆíÀÇ ÃÑ °³¼ö ´Ù. ¿øÇÖÀÎÄÚµù(One-Hot Encoding)°ú ¼ÒÇÁÆ®¸Æ½º(Softmax)ÇÔ¼öPART 5. ±âÃâ ¹× ¿¹»ó ÅäÇÈ108.GPU(Graphic Processing Unit)¿Í CPU(Central Processing Unit)ÀÇ Â÷ÀÌÁ¡109. ¸Ó½Å·¯´× ¸ðµ¨Àº ÇнÀ°ú ÇÔ²² °ËÁõ ¹× Æò°¡ °úÁ¤ÀÌ ÇÊ¿äÇÏ´Ù °¡. ±³Â÷°ËÁõ(k-fold Cross Validation)±â¹ý¿¡ ´ëÇØ ¼³¸íÇϽÿÀ ³ª. ¸Ó½Å·¯´× ¸ðµ¨ÀÇ Æò°¡¹æ¹ý¿¡ ´ëÇÏ¿© ¼³¸íÇϽÿÀ110. ¸Ó½Å·¯´× º¸¾È Ãë¾àÁ¡¿¡ ´ëÇØ ¼³¸íÇϽÿÀ. °¡. ¸Ó½Å·¯´× ÇнÀ°úÁ¤¿¡¼ÀÇ Àû´ëÀû °ø°Ý 4°¡Áö ³ª. °¢°¢ Àû´ëÀû °ø°ÝÀÇ ¹æ¾î ±â¹ý111. µ¥ÀÌÅÍ ¾î³ëÅ×À̼Ç(Data Annotation)112. AIaaS(AI as a Service)¿Í µµÀԽà °í·Á»çÇ×113. ÀüÀÌ ÇнÀ(Transfer Learning)114. Pre-Crime115. Àΰø½Å°æ¸ÁÀÇ ¿À·ù ¿ªÀüÆÄ(Backpropagation) ¾Ë°í¸®Áò116. ¸Ó½Å·¯´×(Machine learning)ÀÇ ÇнÀ¹æ¹ýÀº Å©°Ô 3°¡Áö[ÁöµµÇнÀ(Supervised Learning), ºñÁöµµ ÇнÀ(Unsupervised Learning), °ÈÇнÀ(Reinforcement Learning)]·Î ºÐ·ùÇÑ´Ù. ÀΰøÁö´É¼ÒÇÁÆ®¿þ¾î °³¹ß ÇÁ·Î¼¼½º¸¦ V ¸ðµ¨ ±âÁØÀ¸·Î µµ½ÄÈÇÏ°í °ü·Ã±â¼úÀÇ Ãֽŵ¿Çâ ¹× ¾ÈÀüÃë¾àÁ¡À» ¼³¸íÇϽÿÀ117. ÀΰøÁö´É °³¹ß°úÁ¤¿¡¼ ÁßÁ¡ÀûÀ¸·Î Á¡°ËÇÒ Ç׸ñ118. ÀΰøÁö´É µ¥ÀÌÅÍ Ç°Áú ¿ä±¸»çÇ×119. ¸óÅ×Ä«¸¦·Î(Monte Carlo) Æ®¸®(Tree) Ž»ö(MCTS)120. µðÁöÅÐ Ä«¸£ÅÚ(Digital Cartel)121. XAI(eXplainable AI)122. ÀΰøÁö´É(AI) µ¥ÀÌÅÍ Æò°¡¸¦ À§ÇÑ °í·Á»çÇ×