銆怚PN:matplotlib搴撴渶鏂拌繘灞曘�戙�奡tate of the Library: matplotlib銆媌y Thomas A CaswellO缃戦〉閾炬帴
銆愯鏂�:(Microsoft)涓囦嚎绾у浘璁＄畻骞冲彴GraM銆戙�奊raM: Scaling Graph Computation to the Trillions銆婱 Wu, F Yang, J Xue... (SoCC2015)O缃戦〉閾炬帴
銆愬够鐏�:(Pygotham 2015)鍗旵PU澶ц妯�(鏍稿)闈炵嚎鎬у涔犮�戙�奓arge scale non-linear learning on a single CPU銆媌y Andreas MuellerO缃戦〉閾炬帴浜�:O缃戦〉閾炬帴
銆愬紑婧�:SVM闆嗘垚瀛︿範搴揈nsembleSVM銆�O缃戦〉閾炬帴GitHub:O缃戦〉閾炬帴浠嬬粛鏂囩珷銆奅nsembleSVM: A Library for Ensemble Learning Using Support Vector Machines銆�O缃戦〉閾炬帴
銆�(Python)鏈哄櫒瀛︿範鍥惧儚璇嗗埆瀹炰緥銆戙�奍mage Recognition using Machine Learning Techniques銆媌y prafulkO缃戦〉閾炬帴
銆�"CS224U: Natural Language Understanding"鏂潶绂廚LU2015璇剧▼銆戯紝浠ｇ爜+鏁版嵁銆傚垎甯冨紡璇嶈〃绀� 鍏崇郴鎻愬彇 璇箟parsing 绁炵粡缃戠粶鐢ㄤ簬鑷劧璇█鐞嗚В銆備紬澶氬伐鍏峰拰鎵╁睍闃呰銆�O缃戦〉閾炬帴
銆愬熀浜庢瀬灏忓寲鏋佸ぇ(Minimax)绠楁硶鐨�"unbeatable" Tic Tac Toe銆戙�奣ic Tac Toe: Understanding The Minimax Algorithm銆媌y Jason FoxO缃戦〉閾炬帴
銆怗oogle璇煶杞綍鑳屽悗鐨勭缁忕綉缁溿�戙�奣he neural networks behind Google Voice transcription銆媌y Fran莽oise BeaufaysO缃戦〉閾炬帴pdf:O缃戦〉閾炬帴聽 鎻愪緵鐨勮瘧鏂囥�婅胺姝岃闊宠浆褰曡儗鍚庣殑绁炵粡缃戠粶銆�O缃戦〉閾炬帴
銆愬崥澹鏂�:(David Blei)鏂囨湰/鍥惧儚姒傜巼妯″瀷銆戙�奝robabilistic Models Of Text And Images銆婦avid Meir Blei (2004)O缃戦〉閾炬帴
銆愬崥澹鏂�:鍥�/缁熻寤烘ā鍑镐紭鍖栨柟娉曘�戙�奀onvex Optimization Methods for Graphs and Statistical Modeling銆媀enkat Chandrasekaran (2011)O缃戦〉閾炬帴
銆愬崥澹鏂�:鍙樺垎杩戜技璐濆彾鏂帹鐞嗙畻娉曘�戙�奦ariational Algorithms For Approximate Bayesian Inference銆婱atthew J. Beal (2003)O缃戦〉閾炬帴
銆怭RML銆戙�愯祫鏂欑瑪璁般��//@鐖卞彲鍙�-鐖辩敓娲�:@52nlp缃戠珯鐨�"PRML璇讳功浼�"绯诲垪璧勬枡锛�O缃戦〉閾炬帴鎴栬��@Nietzsche_澶嶆潅缃戠粶鏈哄櫒瀛︿範鐨�"PRML璇讳功浼�"绯诲垪寰崥鏂囩珷:O缃戦〉閾炬帴銆愬嚑涓増鏈殑PRML绗旇銆慗ian Xiao鐨勩�奛otes on Pattern Recognition and Machine Learning (Bishop)銆�O缃戦〉閾炬帴Yuandong Tian鐨勩�奡ome notes on Pattern Recognition and Machine Learning銆�O缃戦〉閾炬帴ChillyRain鐨�"PRML Notes"绯诲垪鍗氭枃:O缃戦〉閾炬帴
闆嗘櫤骞翠細18浣嶄富瑕佹紨璁茶�匬PT鍏紑鍙戝竷锛佸叏鏄共璐э紒娆㈣繋涓嬭浇 鎼溅鎺ㄤ竴涓嬮泦鏅哄勾浼�2015鍙備細鑰呭悇浣嶇殑ppt锛屾秹鍙婁汉宸ユ櫤鑳姐�佹繁搴﹀涔犮�佽绠楃ぞ浼氱瀛︺�佺墿鐞嗗澶氫釜鏂归潰銆�O缃戦〉閾炬帴good book
PRML 聽鐪嬭繃寰堝SVM鐨勮祫鏂欙紝閮芥病鏈夎繖鏈功閲岃竟鐨勫畬鏁寸畝娲佹槑浜�O缃戦〉閾炬帴聽 姣斾竴甯悶宸ョ▼鐨勫啓鐨勬墍璋揝VM瀛︿範蹇冨緱褰撶劧鏄ソ鐨勫锛屼絾鏄噷闈㈡贩娣嗘蹇电殑鍦版柟涔熷緢澶氾紝鎴戜竴搴︿互涓烘垜浼樺寲閲屽鐨凩agrange鍜孧L瀛︾殑涓嶆槸鍚屼竴涓紝鍗曡SVM杩欎竴绔狅紝鍥句功棣嗛噷闅忎究鎵炬湰鍙玈VM鐨勪功閮芥瘮杩欒鐨勭粏銆偮�
闅愰┈灏旂澶ā鍨婬MM(涓�)锛� 鐪嬩簡鏃犳暟鐨勮鏂囷紝閫涗簡鏃犳暟鐨勫潧瀛愪篃娌℃妸闅愰┈寮勬槑鐧斤紝缁堜簬鍦↗uly EDU鐨勮棰戜腑鍜岄偣鍗氱殑璁茶В涓嬬鎳備簡銆�O缃戦〉閾炬帴
Fast Differentially Private Matrix Factorization[Liu,RecSys15]閫氳繃闅忔満姊害鏈椾箣涓囧姩鍔涘鑱旂郴寰垎闅愮鍒拌礉鍙舵柉鍚庨獙閲囨牱O缃戦〉閾炬帴
銆�"鍌荤摐鐗�"Logistic鍥炲綊浠嬬粛銆戙�奓ogistic Regression (for dummies)銆�O缃戦〉閾炬帴
銆怉ndrew Ng鏈哄櫒瀛︿範璇剧▼瀛︿範绗旇銆慗erryLead鎬荤粨鐗�:O缃戦〉閾炬帴浜�:O缃戦〉閾炬帴
銆�(R)闈㈠悜娣卞害瀛︿範鐨勭嚎鎬т唬鏁伴�熸煡銆戙�奓inear Algebra for Deep Learning in R銆媌y Naimish AgarwalO缃戦〉閾炬帴
銆愬崥澹鏂�:(Ilya Sutskever)RNN璁粌銆戙�奣raining Recurrent Neural Networks銆婭lya Sutskever, University of Toronto (2012)O缃戦〉閾炬帴浜�:O缃戦〉閾炬帴
銆愯鏂�+浠ｇ爜:LSTM缃戠粶缁煎悎璇勬祴銆戙�夿enchmarking of LSTM Networks銆婽homas M. Breuel (2015)O缃戦〉閾炬帴GitHub:O缃戦〉閾炬帴
銆愯棰�+璁蹭箟:nVIDIA娣卞害瀛︿範璇剧▼銆�"Deep Learning Courses"O缃戦〉閾炬帴浜�:O缃戦〉閾炬帴
缃戜笂鏈変釜钖涘紑瀹囩殑caffe瀛︿範绗旇锛岃浆涓瘮杈冨叏鐨勯泦鎴愮増pdf:O缃戦〉閾炬帴
銆�(Python)NN璁粌杩囩▼鍙鍖�(闄勬簮鐮�)銆戙�奦ideo of a neural network learning銆媌y Go to the profile of Milo Spencer-HarperO缃戦〉閾炬帴GitHub(Neural Network Animation):
銆�(R)澶氬眰鍒嗙被鍙橀噺Impact Coding鍥炲綊寤烘ā銆戙�奙odeling Trick: Impact Coding Of Categorical Variables With Many Levels銆媌y Nina ZumelO缃戦〉閾炬帴
銆愮粰寮�鍙戣�呯殑鏈哄櫒瀛︿範璺佃鎸囧崡銆戙�奙achine Learning for Programmers: Leap from developer to machine learning practitioner銆媌y Jason BrownleeO缃戦〉閾炬帴
銆愬紑婧�:鐥呮�佺煩闃电殑杩唬姹傞�咺nverseProblem銆�"This function inverts ill conditioned matrices using an iterative solution to the Tikhonov regularization problem"O缃戦〉閾炬帴
[IPN] Kalman Filter textbook using Ipython NotebookO缃戦〉閾炬帴鍗″皵鏇兼护娉㈠櫒鏁欐潗锛岀敤灏介噺灏戠殑鏁板鍜屾帹瀵硷紝浼犳巿鐩磋鍜岀粡楠岋紝鍏ㄩ儴Python绀轰緥锛屽唴瀹硅鐩栧崱灏旀浖婊ゆ尝鍣ㄣ�佹墿灞曞崱灏旀浖婊ゆ尝锛屾棤杩瑰崱灏旀浖婊ゆ尝绛夛紝鍖呮嫭缁冧範鍜屽弬鑰冪瓟妗堬紝鎻愪緵PDF鐗� ipn:O缃戦〉閾炬帴浜�:O缃戦〉閾炬帴銆奒alman and Bayesian Filters in Python銆嬫渶鏂扮増(2015.8.9):O缃戦〉閾炬帴
銆愯棰�:缁撴瀯娣卞害瀛︿範銆戙�奃eep Learning with Structure, Charlie Tang, Uni of Toronto - RE.WORK Deep Learning Summit 2015銆�O缃戦〉閾炬帴浜�:O缃戦〉閾炬帴鍙傞槄:O鐖卞彲鍙�-鐖辩敓娲�
銆怣achine Learning Group,University of Cambridge鐨凱ublications銆�O缃戦〉閾炬帴
銆愬崥澹鏂�:闈㈠悜璇煶璇嗗埆/璁＄畻鍖栧/鑷劧璇█澶勭悊鐨勬繁搴﹀涔犳柟娉曘�戙�奃eep learning approaches to problems in speech recognition, computational chemistry, and natural language text processing銆婫eorge E. Dahl, University of Toronto (2015)O缃戦〉閾炬帴浜�:O缃戦〉閾炬帴
銆愮澹鏂�:绁炵粡缃戠粶Dropout浼樺寲銆戙�奍mproving Neural Networks with Dropout銆婲itish Srivastava,University of Toronto (2013)O缃戦〉閾炬帴浜�:O缃戦〉閾炬帴
銆怉wesome闆嗗悎澶у叏銆戣鐩栧钩鍙般�佺紪绋嬭瑷�銆佸墠绔紑鍙戙�佸悗绔紑鍙戙�佽绠楁満绉戝銆佸ぇ鏁版嵁銆佺悊璁恒�佷功绫嶃�佺紪杈戝櫒銆佹父鎴忋�佸紑鍙戠幆澧冦�佸ū涔愩�佹暟鎹簱銆佽祫婧愩�佸畨鍏ㄧ瓑涓婚锛屽牚绉癆wesome涔婣wesome GitHub:O缃戦〉閾炬帴pdf:O缃戦〉閾炬帴鍙傞槄O鐖卞彲鍙�-鐖辩敓娲�
銆愬紑婧�:(Python)Random Bits Regression+FTRL=Randomly Follow the Regularized Leader鍦ㄧ嚎瀛︿範鍒嗙被鍣ㄣ��"Online Random Bit Regression with FTRL-Proximal in Python"O缃戦〉閾炬帴
銆怐eep Generative Models銆慏eep Generative ModelsO缃戦〉閾炬帴Bengio浜�8鏈�12鏃ュ湪DL鏆戞湡鐝粙缁嶆繁搴︾敓鎴愭ā鍨嬶紝绗�15椤靛紑璁睤enoising AE锛岀29椤垫彁鍒癏elmholtz machine鍜孷ariational AE锛岀粨璁虹洰鍓嶆渶濂界殑鍥惧儚妯″瀷鏄� Generative Adversarial NetsO缃戦〉閾炬帴
鐧惧害浣欏嚡鍥㈤槦鏈�鏂板姏浣溿�婂熀浜庢繁搴﹀涔犵殑鍥惧儚璇嗗埆杩涘睍锛氱櫨搴︾殑鑻ュ共瀹炶返銆嬶紝閲岄潰涓嶅皯鏈変环鍊肩殑瀹炶返缁忛獙銆備綔鑰咃細閮藉ぇ榫欍�佷綑杞跺崡銆佺綏鎭掞紝鍏朵粬浣滆�咃細寮犲仴銆侀粍鐣呫�佸緪浼熴�佷綑鍑�傘�婁腑鍥借绠楁満瀛︿細閫氳銆嬬11鍗风4鏈燂紝2015骞�4鏈堛�傛垜瀛︿範浜嗕竴涓嬶紝鍐欎簡涓憳瑕佸拰涓汉浣撲細锛屼粎渚涘弬鑰冿紝鏂囩尞鍙笅杞姐��O缃戦〉閾炬帴
Michael Collins鑷劧璇█澶勭悊涓殑鏈哄櫒瀛︿範娉� 1) MIT03骞淬�侼LP浠诲姟: 璇嶆�ф爣娉� 鍙ユ硶鍒嗘瀽 WSD IE MT锛汳L鏂规硶: PCFG HMM 浼拌鍜屽钩婊戞妧鏈� EM绠楁硶 瀵规暟绾挎�фā鍨� 鏍告柟娉�O缃戦〉閾炬帴2) 鍝ヤ鸡姣曚笟12骞淬��O缃戦〉閾炬帴3) coursera13骞淬��O缃戦〉閾炬帴4锛夊熀鏈病鏈夌缁忕綉缁滃拰娣卞害瀛︿範娉�?
銆愭棤鐩戠潱銆戙�怢STM瀛︿範銆戣棰戣〃绀�<缂栫爜,瑙ｇ爜>LSTM瀛︿範瑙嗛琛ㄧず Unsupervised Learning of Video Representations using LSTMs[Srivastava,ICML15] 鐢ㄤ竴涓紪鐮丩STM灏嗚棰戝簭鍒楁槧灏勪负鍥哄畾闀垮害鐨勮〃绀猴紝鐢ㄤ竴涓�/澶氫釜LSTM瑙ｇ爜璇ヨ〃绀哄彲鎵ц涓嶅悓鐨勪换鍔★紝濡傞噸鏋勮緭鍏ュ簭鍒楁垨鑰呴娴嬫湭鐭ュ簭鍒椼��O缃戦〉閾炬帴鏃犵洃鐫ｅ涔犳彁鍗囨湁鐩戠潱娲诲姩璇嗗埆浠诲姟
銆愭満鍣ㄥ涔犵畻娉曟瘮杈冦�戝湪涓嶅悓璇█鎴栧钩鍙颁笅鐨勬椂闂村拰鎬ц兘鍒嗘瀽姣旇緝闈炲父鏈夋剰鎬濈殑涓�涓瘮杈冿細涓嶅悓鏈哄櫒瀛︿範绠楁硶鍦ㄤ笉鍚岃瑷�鎴栧钩鍙颁笅鐨勬椂闂村拰鎬ц兘鍒嗘瀽姣旇緝锛宐enchmark for scalability/speed and accuracy of machine learning libraries for classification锛岃锛�O缃戦〉閾炬帴
鍓嶄簺鏃ュ瓙璋锋瓕鐨勫崥瀹�O缃戦〉閾炬帴浠嬬粛浜嗏�滈�犳ⅵ(inceptionism)鈥濈殑鏂瑰紡鍙鍖朑oogLeNet classes銆傚悗鏉ュ張濂戒簨鑰呯湡鐨勬湁鏍峰鏍锋悶鍑烘潵浜�O缃戦〉閾炬帴锛屾簮浠ｇ爜鍦ㄨ繖閲�O缃戦〉閾炬帴
銆恜ython鑿滈笩鏃ヨ7銆�O缃戦〉閾炬帴锛堝垎浜嚜@涓�鐐硅祫璁�锛�
銆愮粨鏋勬渶澶х喌妯″瀷銆戠粨鏋勬渶澶х喌妯″瀷SME锛屾彁鍗嘗1姝ｅ垯鐨勬渶澶х喌绠楁硶 Structural Maxent Models [Cortes,ICML15] 鏉′欢SME/澶氶」LR锛岃仈绯讳簩绫�/澶氱被娣卞害Boosting绠楁硶O缃戦〉閾炬帴鏈�澶х喌妯″瀷寰堟柟渚跨殑鍔犲叆鍚勭鐗瑰緛锛屽杈撳叆鍙橀噺鐨勯槇鍊煎嚱鏁般�傗�滅粨鏋勨�濇寚浠庡鏉傛�т笉鏂鍔犵殑鍑芥暟鏃忎腑閫夋嫨鐗瑰緛锛屼絾鏇村鏉傛ā鍨嬬殑鎬诲弬鏁版潈閲嶈閫掑噺
銆愭椂闂村簭鍒椼�戙�愮浉浼奸噸澶嶇粨鏋勩�戙�愬紓甯告娴嬨�戙�愬够鐏�:鏃跺簭閲嶅缁撴瀯(妯″紡)鎸栨帢鈥斺�旂畻娉曚笌搴旂敤銆戙�奆inding Repeated Structure in Time Series: Algorithms and Applications銆媌y Abdullah Mueen, Eamonn Keogh (SDM2015 Tutorial)O缃戦〉閾炬帴
銆愭繁搴﹂珮鏂繃绋�(GP)銆戙�愮煡璇嗚縼绉诲涔犮�戞繁搴﹂珮鏂繃绋�(GP)鐭ヨ瘑杩佺Щ瀛︿範 Asymmetric Transfer Learning with Deep Gaussian Processes [Kandemir,ICML15]灏嗙洰鏍囧煙鏁版嵁閫氳繃婧愬煙鐨勭涓�灞侴P鎶曞奖鍒版簮鍩熺殑闅愯棌绌洪棿, 灏嗙洰鏍囧煙鏁版嵁閫氳繃鐩爣鍩熺殑绗竴灞侴P鎶曞奖鍒扮洰鏍囧煙鐨勯殣钘忕┖闂�; 鐒跺悗绾挎�х粍鍚堣繖涓ょ琛ㄧず,鍠傜粰鐩爣鍩熺殑绗簩灞侴PO缃戦〉閾炬帴
銆愯縼绉诲涔犮�戙�愭繁搴﹀涔犮�戙�愯縼绉诲涔犲疄鐜版繁搴﹀涔犳ā鍨嬪啀鍒╃敤銆戙�奟ecycling Deep Learning Models with Transfer Learning銆媌y Zachary Chase LiptonO缃戦〉閾炬帴
銆愯绋�:(Dataquest.io)鐢≒ython鍋氬晢涓�(鏁版嵁)鍒嗘瀽銆戙�奝ython for Business Analysts - Learn how Python can supercharge your data analysis workflow銆�O缃戦〉閾炬帴
銆愭姂鍒跺浘鍍弒peckle鍣０鐨凪ultilook鎶�鏈�戙�奙ultilook Technique for speckle reduction銆�O缃戦〉閾炬帴
銆怭ython鏍稿Dataframes:Dask/OpenStreetMap銆戙�奜ut-of-Core Dataframes in Python: Dask and OpenStreetMap銆媌y Jake VanderplasO缃戦〉閾炬帴
銆愬啓缁橰鑿滈笩鐨刧gplot2鏁版嵁鍙鍖栨寚鍗椼�戙�奍ntroduction to ggplot2 for people who don't know R銆�O缃戦〉閾炬帴
銆愮墰娲ュぇ瀛︽繁搴﹀涔犺绋�(2015)銆戙�奃eep learning at Oxford 2015銆�O缃戦〉閾炬帴Youtube:O缃戦〉閾炬帴鐗涙触澶уNando de Freitas涓昏鐨勬満鍣ㄥ涔犺绋嬶紝閲嶇偣浠嬬粛娣卞害瀛︿範锛岃繕璇锋潵Deepmind鐨凙lex Graves鍜孠arol Gregor瀹㈠骇鎶ュ憡锛屽唴瀹广�佽瑙ｉ兘灞炰竴娴侊紝寮虹儓鎺ㄨ崘锛� 浜�:O缃戦〉閾炬帴聽 Course materials(Lectures+Practicals):O缃戦〉閾炬帴GitHub:O缃戦〉閾炬帴
銆愬够鐏�:闈㈠悜缃戠粶鎼滅储鍜岃嚜鐒惰瑷�澶勭悊鐨勬繁搴﹀涔�(DSSM/RNN)銆戙�奃eep Learning for Web Search and Natural Language Processing銆媌y Jianfeng Gao [Microsoft] (WSDM 2015)O缃戦〉閾炬帴浜�:O缃戦〉閾炬帴
銆愯绋�:(Dataquest.io)鐢≒ython鍋氬晢涓�(鏁版嵁)鍒嗘瀽銆戙�奝ython for Business Analysts - Learn how Python can supercharge your data analysis workflow銆�O缃戦〉閾炬帴
銆怋ayesian optimisation for smart hyperparameter search銆� - Tim HeadO缃戦〉閾炬帴
銆愯鏂�:闈㈠悜Spark澶ф暟鎹帹鑽愮殑鍗忓悓杩囨护骞惰鍔犻�熺畻娉旳LS-NCG銆戙�夾lgorithmic Acceleration of Parallel ALS for Collaborative Filtering: Speeding up Distributed Big Data Recommendation in Spark銆婱 Winlaw, M Hynes, A Caterini, H Sterck (2015)O缃戦〉閾炬帴
銆愯棰�+骞荤伅:Spark/GraphX澶ц妯″浘鍒嗘瀽浼樺寲缁忛獙鍒嗕韩銆戙�奅xperience and Lessons Learned for Large-Scale Graph Analysis using GraphX銆媌y Jason, DaiO缃戦〉閾炬帴浜�:O缃戦〉閾炬帴
銆愭暀绋嬨�戙�愭櫘鏋楁柉椤裤�戙�愬箍涔夌嚎鎬фā鍨嬨�慙ecture Notes on Generalized Linear Models [Rodr铆guez, 07]. 浜屽厓鍝嶅簲鐨凩ogit妯″瀷锛坙ogistic锛宲robit锛夈�傝鏁般�佸垪鑱旇〃銆佸瓨娲绘暟鎹殑娉婃澗妯″瀷銆傚鍏冨搷搴旂殑澶氶」LR 銆�O缃戦〉閾炬帴闄や簡PDF鐗堝杩樻湁MathJax缂栬緫鐨勭綉椤电増
銆愪唬鐮佸伐鍏峰钩鍙般�戙�愬崱鍐呭熀姊呴殕瀹為獙瀹ゃ�戝崱鍐呭熀姊呴殕SELECT瀹為獙瀹ゅ紑婧愪唬鐮�. 1)骞惰鍧愭爣涓嬮檷瑙1姝ｅ垯椋庨櫓鏈�灏忓寲: 骞惰LASSO鍜岀█鐤廘R [ICML11] 2)鍩轰簬GraphLab鐨勫苟琛孏ibbs閲囨牱 [AIStats11] 3)GraphLab:骞惰鏈哄櫒瀛︿範妗嗘灦 [UAI10] 4)骞惰椹皵绉戝か闅忔満鍦篗RF鎺ㄧ悊 鏍戞潯浠堕殢鏈哄満CRF缁撴瀯瀛︿範 鍒嗗竷寮忓洜瀛愬浘鎺ㄧ悊 瀛愭ā鍑芥暟浼樺寲O缃戦〉閾炬帴
Memory, Reading, and Comprehension (pdf)O缃戦〉閾炬帴Phil Blunsom 8鏈�10鏃ュ湪Deep Learning Summer School鐨勮搴с�傝鍒板悗闈㈠熀鏈氨鏄笅闈㈤偅绡囪鏂囦簡 銆愯鏂�:璇█缈昏瘧鏂归潰鍗曞眰RNN+Memory浼樹簬鏇存繁灞傜綉缁溿�戙�奓earning to Transduce with Unbounded Memory銆婨 Grefenstette, KM Hermann, M Suleyman, P Blunsom [DeepMind] (2015)O缃戦〉閾炬帴
銆愭暀绋嬨�戙�愮缁忕綉缁溿�戙�奣he Nature of Code銆嬶紝Chapter 10. Neural NetworksO缃戦〉閾炬帴
铻嶅悎澶氬厓淇℃伅鐨勬贩鍚堟帹鑽�: 鐢ㄦ埛闂寸浉浼� 浜у搧闂寸浉浼� 鐢ㄦ埛浜у搧闂存墦鍒�. HyPER: A Flexible and Extensible Probabilistic Framework for Hybrid Recommender Systems [Kouki,RecSys15] 娣峰悎鎺ㄨ崘褰㈠紡鍖栦负閾伴摼鎹熷け鐨勯┈鍏嬬澶殢鏈哄満,鐢ㄦ鐜囩紪绋嬭瑷�probabilistic soft logic鏋勫缓鎺ㄨ崘绯荤粺.O缃戦〉閾炬帴
銆奝ython 涓嶆槸 C銆嬫垜涓�鐩翠娇鐢� Python锛岀敤瀹冨鐞嗗悇绉嶆暟鎹瀛﹂」鐩�� Python 浠ユ槗鐢ㄩ椈鍚嶃�傛湁缂栫爜缁忛獙鑰呭涔犳暟澶╁氨鑳戒笂鎵嬶紙鎴栨湁鏁堜娇鐢ㄥ畠锛夈��O缃戦〉閾炬帴锛坥schina 璇戯級
銆愯鏂�+浠ｇ爜+鏁版嵁:闈㈠悜鎯呮劅鏍囩鍒嗗竷棰勬祴鐨勫崐鐩戠潱閫掑綊鑷姩缂栫爜鍣ㄣ�戙�奡emi-Supervised Recursive Autoencoders for Predicting Sentiment Distributions銆� R Socher, J Pennington, E Huang, Andrew Y. Ng, C Manning (EMNLP 2011)O缃戦〉閾炬帴page(code+data):O缃戦〉閾炬帴
銆愬崥澹鏂�:闈㈠悜鑷劧璇█澶勭悊鍜屾満鍣ㄨ瑙夌殑閫掑綊娣卞害瀛︿範銆戙�奟ecursive Deep Learning for Natural Language Processing and Computer Vision銆媌y Richard Socher (Stanford 2014) "2014 Arthur L. Samuel Best Computer Science PhD Thesis Award"O缃戦〉閾炬帴浜�:O缃戦〉閾炬帴
銆愯鏂�+浠ｇ爜(Python/Theano):CNN鍙ュ瓙鍒嗙被銆戙�奀onvolutional Neural Networks for Sentence Classification銆媃oon Kim (EMNLP 2014)O缃戦〉閾炬帴GitHub:O缃戦〉閾炬帴GitXiv:O缃戦〉閾炬帴