人工智能詞匯117

常用英語詞匯-andrewNg課程trainingset訓練集合intensity強度Regression回歸Lossfunction損失函數(shù)non-convex非凸函數(shù)neuralnetwork神經(jīng)網(wǎng)絡supervisedlearning監(jiān)督學習regressionproblem回歸問題處理的是連續(xù)的問題classificationproblem分類問題discreetvalue離散值supportvectormachines支持向量機learningtheory學習理論learningalgorithms學習算法unsupervisedlearning無監(jiān)督學習gradientdescent梯度下降linearregression線性回歸NeuralNetwork神經(jīng)網(wǎng)絡gradientdescent梯度下降normalequationslinearalgebra線性代數(shù)superscript上標trainingexample訓練樣本hypothesis假設，用來表示學習算法的輸出LMSalgorithm“l(fā)eastmeansquares最小二乘法算法batchgradientdescent批量梯度下降constantlygradientdescent隨機梯度下降iterativealgorithm迭代算法partialderivative偏導數(shù)contour等高線quadraticfunction二元函數(shù)locallyweightedregression局部加權回歸underfitting欠擬合overfitting過擬合non-parametriclearningalgorithms無參數(shù)學習算法parametriclearningalgorithm參數(shù)學習算法activation激活值activationfunction激活函數(shù)additivenoise加性噪聲autoencoder自編碼器Autoencoders自編碼算法exponentiationexponentiation指數(shù)averagefiringrate平均激活率averagesum-of-squareserroraveragesum-of-squareserror均方差backpropagation后向傳播basis基basisfeaturevectors特征基向量batchgradientascent批量梯度上升法Bayesianregularizationmethod貝葉斯規(guī)則化方法Bernoullirandomvariable伯努利隨機變量biasterm偏置項binaryclassfication二元分類classlabels類型標記concatenation級聯(lián)conjugategradient共軛梯度contiguousgroups聯(lián)通區(qū)域convexoptimizationsoftware凸優(yōu)化軟件convolution卷積costfunction代價函數(shù)covariancematrix協(xié)方差矩陣DCcomponent直流分量decorrelation去相關degeneracy退化demensionalityreduction降維diagonal對角線diffusionofgradients梯度的彌散eigenvalue特征值eigenvector特征向量errorterm殘差featurematrix特征矩陣featurestandardization特征標準化feedforwardarchitectures前饋結(jié)構(gòu)算法feedforwardneuralnetwork前饋神經(jīng)網(wǎng)絡feedforwardpass前饋傳導fine-tuned微調(diào)first-orderfeature一階特征forwardpass前向傳導forwardpropagation前向傳播Gaussianprior高斯先驗概率generativemodel生成模型gradientdescent梯度下降Greedylayer-wisetraining逐層貪婪訓練方法groupingmatrix分組矩陣Hadamardproduct阿達馬乘積HessianmatrixHessian矩陣hiddenlayer隱含層derivative導函數(shù)hiddenunits隱藏神經(jīng)元linearsuperpositionhiddenunits隱藏神經(jīng)元Hierarchicalgrouping層次型分組higher-orderfeatures更高階特征highlynon-convexoptimizationproblem高度非凸的優(yōu)化問題histogram直方圖hyperbolictangent雙曲正切函數(shù)hypothesis估值，假設identityactivationfunction恒等激勵函數(shù)IID獨立同分布illumination照明inactive抑制independentcomponentanalysis獨立成份分析inputdomains輸入域inputlayer輸入層intensity亮度/灰度interceptterm截距KLdivergence相對熵KLdivergenceKL分散度k-MeansK-均值learningrate學習速率leastsquares最小二乘法line-searchalgorithm線搜索算法localmeansubtraction局部均值消減localoptima局部最優(yōu)解logisticregression邏輯回歸lossfunction損失函數(shù)low-passfiltering低通濾波magnitude幅值MAP極大后驗估計maximumlikelihoodestimation極大似然估計mean平均值MFCCMel倒頻系數(shù)multi-classclassification多元分類neuralnetworks神經(jīng)網(wǎng)絡neuron神經(jīng)元Newton’smethod牛頓法non-convexfunction非凸函數(shù)non-linearfeature非線性特征norm范式normbounded有界范數(shù)normconstrained范數(shù)約束normalization歸一化linearcorrespondence線性響應reconstructioncost重建代價numericalroundofferrorsreconstructioncost重建代價numericallychecking數(shù)值檢驗numericallyreliable數(shù)值計算上穩(wěn)定objectdetection物體檢測objectivefunction目標函數(shù)off-by-oneerror缺位錯誤orthogonalization正交化outputlayer輸出層overallcostfunction總體代價函數(shù)over-completebasis超完備基over-fitting過擬合partsofobjects目標的部件part-wholedecompostion部分-整體分解PCA主元分析penaltyterm懲罰因子per-examplemeansubtraction逐樣本均值消減pooling池化pretrain預訓練principalcomponentsanalysis主成份分析quadraticconstraints二次約束RBMs受限Boltzman機reconstructionterm重構(gòu)項redundant冗余reflectionmatrix反射矩陣regularization正則化regularizationterm正則化項rescaling縮放robust魯棒性run行程second-orderfeature二階特征sigmoidactivationfunctionS型激勵函數(shù)significantdigits有效數(shù)字singularvalue奇異值singularvector奇異向量smoothedL1penalty平滑的L1范數(shù)懲罰SmoothedtopographicL1sparsitypenalty平滑地形L1稀疏懲罰函數(shù)smoothing平滑SoftmaxRegressonSoftmax回歸sortedindecreasingorder降序排列sourcefeatures源特征sparseautoencodersparseautoencoder消減歸一化reconstructionbasedmodels基于重構(gòu)的模型transformationtransformation變換translationinvariant平移不變性trivialanswer平凡解under-completebasis不完備基unrolling組合擴展unsupervisedlearning無監(jiān)督學習variance方差vecotrizedimplementation向量化實現(xiàn)vectorization矢量化visualcortex視覺皮層weightdecay權重衰減weightedaverage加權平均值whitening白化zero-mean均值為零Accumulatederrorbackpropagation累積誤差逆?zhèn)鞑ctivationFunction激活函數(shù)AdaptiveResonanceTheory/ART自適應諧振理論Addictivemodel加性學習AdversarialNetworks對抗網(wǎng)絡AffineLayer仿射層Affinitymatrix親和矩陣Agent代理/智能體Sparsity稀疏性sparsityparameter稀疏性參數(shù)sparsitypenalty稀疏懲罰squarefunction平方函數(shù)squared-error方差stationary平穩(wěn)性（不變性）stationarystochasticprocess平穩(wěn)隨機過程step-size步長值supervisedlearning監(jiān)督學習symmetricpositivesemi-definitematrix對稱半正定矩陣symmetrybreaking對稱失效tanhfunction雙曲正切函數(shù)theaverageactivation平均活躍度thederivativecheckingmethod梯度驗證方法theempiricaldistribution經(jīng)驗分布函數(shù)theenergyfunction能量函數(shù)theLagrangedual拉格朗日對偶函數(shù)theloglikelihood對數(shù)似然函數(shù)thepixelintensityvalue像素灰度值therateofconvergence收斂速度topographiccostterm拓撲代價項topographicordered拓撲秩序AlgorithmAlgorithm算法BatchNormalization/BN批量歸一化Baselearningalgorithm基學習算法Baselearningalgorithm基學習算法Class-conditionalprobability類條件概率Alpha-betapruninga-0剪枝Anomalydetection異常檢測Approximation近似AreaUnderROCCurve／AUCRoc曲線下面積ArtificialGeneralIntelligence/AGI通用人工智能ArtificialIntelligence/AI人工智能Associationanalysis關聯(lián)分析Attentionmechanism注意力機制Attributeconditionalindependenceassumption屬性條件獨立性假設Attributespace屬性空間Attributevalue屬性值Autoencoder自編碼器Automaticspeechrecognition自動語音識別Automaticsummarization自動摘要Averagegradient平均梯度Average-Pooling平均池化BackpropagationThroughTime通過時間的反向傳播Backpropagation/BP反向傳播Baselearner基學習器Bayesdecisionrule貝葉斯判定準則BayesModelAveraging／BMA貝葉斯模型平均Bayesoptimalclassifier貝葉斯最優(yōu)分類器Bayesiandecisiontheory貝葉斯決策論Bayesiannetwork貝葉斯網(wǎng)絡Between-classscattermatrix類間散度矩陣Bias偏置/偏差Bias-variancedecomposition偏差-方差分解Bias-VarianceDilemma偏差-方差困境Bi-directionalLong-ShortTermMemory/Bi-LSTM雙向長短期記憶Binaryclassification二分類Binomialtest二項檢驗Bi-partition二分法Boltzmannmachine玻爾茲曼機Bootstrapsampling自助采樣法／可重復采樣Bootstrapping自助法Break-EventPoint／BEP平衡點Calibration校準Cascade-Correlation級聯(lián)相關Categoricalattribute離散屬性Classificationandregressiontree/CART分類與回ConditionalProbabilityTable／CPT條件概率表Cost-sensitive成本敏感Cost-sensitive成本敏感Conditionalmutualinformation條件互信息歸樹Classifier分類器Class-imbalance類別不平衡Closed-form閉式Cluster簇/類/集群Clusteranalysis聚類分析Clustering聚類Clusteringensemble聚類集成Co-adapting共適應Codingmatrix編碼矩陣COLT國際學習理論會議Committee-basedlearning基于委員會的學習Competitivelearning競爭型學習Componentlearner組件學習器Comprehensibility可解釋性ComputationCost計算成本ComputationalLinguistics計算語言學Computervision計算機視覺Conceptdrift概念漂移ConceptLearningSystem/CLS概念學習系統(tǒng)Conditionalentropy條件熵Conditionalrandomfield/CRF條件隨機場Conditionalrisk條件風險Confidence置信度Confusionmatrix混淆矩陣Connectionweight連接權Connectionism連結(jié)主義Consistency一致性／相合性Contingencytable列聯(lián)表Continuousattribute連續(xù)屬性Convergence收斂Conversationalagent會話智能體Convexquadraticprogramming凸二次規(guī)劃Convexity凸性Convolutionalneuralnetwork/CNN卷積神經(jīng)網(wǎng)絡Co-occurrence同現(xiàn)Correlationcoefficient相關系數(shù)Cosinesimilarity余弦相似度Costcurve成本曲線CostFunction成本函數(shù)Costmatrix成本矩陣Crossentropy交叉熵Crossvalidation交叉驗證Crowdsourcing眾包Curseofdimensionality維數(shù)災難Cutpoint截斷點Cuttingplanealgorithm割平面法Datamining數(shù)據(jù)挖掘Dataset數(shù)據(jù)集DecisionBoundary決策邊界Decisionstump決策樹樁Decisiontree決策樹／判定樹Deduction演繹DeepBeliefNetwork深度信念網(wǎng)絡DeepConvolutionalGenerativeAdversarialNetworkDCGAN深度卷積生成對抗網(wǎng)絡Deeplearning深度學習Deepneuralnetwork/DNN深度神經(jīng)網(wǎng)絡DeepQ-Learning深度Q學習DeepQ-Network深度Q網(wǎng)絡Densityestimation密度估計Density-basedclustering密度聚類Differentiableneuralcomputer可微分神經(jīng)計算機Dimensionalityreductionalgorithm降維算法Directededge有向邊Disagreementmeasure不合度量Discriminativemodel判別模型Discriminator判別器Distancemeasure距離度量Distancemetriclearning距離度量學習Distribution分布Divergence散度Diversitymeasure多樣性度量／差異性度量Domainadaption領域自適應Downsampling下采樣D-separation(Directedseparation)有向分離Dualproblem對偶問題Dummynode啞結(jié)點DynamicFusion動態(tài)融合Dynamicprogramming動態(tài)規(guī)劃Eigenvaluedecomposition特征值分解Embedding嵌入Emotionalanalysis情緒分析Empiricalconditionalentropy經(jīng)驗條件熵

Empiricalentropy經(jīng)驗熵Featureselection特征選擇Empiricalentropy經(jīng)驗熵Featureselection特征選擇Empiricalrisk經(jīng)驗風險End-to-End端到端Energy-basedmodel基于能量的模型Ensemblelearning集成學習Ensemblepruning集成修剪ErrorCorrectingOutputCodes／ECOC糾錯輸出碼Errorrate錯誤率Error-ambiguitydecomposition誤差-分歧分解Euclideandistance歐氏距離Evolutionarycomputation演化計算Expectation-Maximization期望最大化Expectedloss期望損失ExplodingGradientProblem梯度爆炸問題Exponentiallossfunction指數(shù)損失函數(shù)ExtremeLearningMachine/ELM超限學習機Factorization因子分解Falsenegative假負類Falsepositive假正類FalsePositiveRate/FPR假正例率Featurevector特征向量FeaturedLearning特征學習FeedforwardNeuralNetworks/FNN前饋神經(jīng)網(wǎng)絡Fine-tuning微調(diào)Flippingoutput翻轉(zhuǎn)法Fluctuation震蕩Forwardstagewisealgorithm前向分步算法Frequentist頻率主義學派Full-rankmatrix滿秩矩陣Functionalneuron功能神經(jīng)元Gainratio增益率Gametheory博弈論Gaussiankernelfunction高斯核函數(shù)GaussianMixtureModel高斯混合模型GeneralProblemSolving通用問題求解Generalization泛化Generalizationerror泛化誤差Generalizationerrorbound泛化誤差上界GeneralizedLagrangefunction廣義拉格朗日函數(shù)GeneralizedRayleighquotient廣義瑞利商Featureengineering特征工程GeneralizedRayleighquotient廣義瑞利商Featureengineering特征工程GenerativeAdversarialNetworks/GAN生成對抗網(wǎng)絡GenerativeModel生成模型Generator生成器GeneticAlgorithm/GA遺傳算法Gibbssampling吉布斯采樣Giniindex基尼指數(shù)Globalminimum全局最小GlobalOptimization全局優(yōu)化Gradientboosting梯度提升GradientDescent梯度下降Graphtheory圖論Ground-truth真相／真實Hardmargin硬間隔Hardvoting硬投票Harmonicmean調(diào)和平均Hessematrix海塞矩陣Hiddendynamicmodel隱動態(tài)模型Hiddenlayer隱藏層HiddenMarkovModel/HMM隱馬爾可夫模型Hierarchicalclustering層次聚類Hilbertspace希爾伯特空間Hold-out留出法Homogeneous同質(zhì)Hybridcomputing混合計算Hyperparameter超參數(shù)Hypothesis假設Hypothesistest假設驗證ICML國際機器學習會議Improvediterativescaling/IIS改進的迭代尺度法Incrementallearning增量學習Independentandidenticallydistributed/i.i.d.獨立同分布IndependentComponentAnalysis/ICA獨立成分分析Indicatorfunction指示函數(shù)Individuallearner個體學習器Induction歸納Inductivebias歸納偏好Inductivelearning歸納學習InductiveLogicProgramming／ILP歸納邏輯程序設計Informationentropy信息熵Informationgain信息增益Inputlayer輸入層Hingelossfunction合頁損失函數(shù)Insensitiveloss不敏感