1、Letter AAccumulated error backpropagation累積誤差逆?zhèn)鞑ctivation Function 激活函數(shù)Adaptive Resonance Theory/ART自適應諧振理論Addictive model加性學習Adversarial Networks 對抗網(wǎng)絡Affine Layer仿射層Affinity matrix親和矩陣Agent 代理/智能體 Algorithm 算法Alpha-beta pruning a - 3 剪枝 Anomaly detection 異常檢測 Approximation 近似 Area Under ROC Curve

2、/AUC Roc 曲線下面積 Artificial General Intelligence/AGI通用人工智能Artificial Intelligence/AI人工智能Association analysis關(guān)聯(lián)分析Attention mechanism 注意力機制Attribute conditional independence assumption屬性條件獨立性彳民設Attribute space屬性空間Attribute value屬性值Autoencoder自編碼器Automatic speech recognition自 動語音識別Automatic summarization

3、 自動摘要Average gradient 平均梯度Average-Pooling 平均池化 Letter BBackpropagation Through Time 通過時間的反向傳播Backpropagation/BP 反向傳播Base learner基學習器Base learning algorithm基學習算法Batch Normalization/BN批量歸一化Bayes decision rule貝葉斯判定準貝UBayes Model Averaging / BMA 貝葉斯模型平均Bayes optimal classifier貝葉斯最優(yōu)分類器Bayesian decision t

4、heory貝葉斯決策論Bayesian network貝葉斯網(wǎng)絡Between-class scatter matrix類間散度矩陣Bias 偏置/偏差Bias-variance decomposition偏差一方差分解Bias-Variance Dilemma偏差 一方差困境Bi-directional Long-Short Term Memory/Bi-LSTM雙向長短期記憶Binary classification二分類Binomial test二項檢驗Bi-partition二分法Boltzmann machine玻爾茲曼機Bootstrap sampling自助采樣法/可重復采樣/有

5、放回采樣Bootstrapping自助法Break-Event Point / BEP 平衡點Letter CCalibration 校準Cascade-Correlation級聯(lián)相關(guān)Categorical attribute離散屬性Class-conditional probability類條件概率Classification and regression tree/CART分類與回歸樹Classifier 分類器Class-imbalance類別不平衡Closed -form閉式Cluster 簇/類/集群Cluster analysis聚類分析Clustering 聚類Clusterin

6、g ensemble聚類集成Co-adapting 共適應Coding matrix編碼矩陣COLT國際學習理論會議Committee-based learning基于委員會的學習Competitive learning競爭型學習Component learner組件學習器Comprehensibility可解釋性Computation Cost計算成本Computational Linguistics計算語言學Computer vision 計算機視覺Concept drift 概念漂移Concept Learning System /CLS 概念學習系統(tǒng)Conditional entrop

7、y 條件嫡Conditional mutual information條件互信息Conditional Probability Table /CPT 條件概率表Conditional random field/CRF條件隨機場Conditional risk條件風險Confidence 置信度Confusion matrix混淆矩陣Connection weight連接權(quán)Connectionism連結(jié)主義Consistency 一致性/相合性Contingency table歹 U 聯(lián)表Continuous attribute連續(xù)屬性Convergence 收斂Conversational a

8、gent會話智能體Convex quadratic programming凸二次規(guī)戈 ijConvexity 凸性Convolutional neural network/CNN卷積神經(jīng)網(wǎng)絡Co-occurrence 同現(xiàn)Correlation coefficient相關(guān)系數(shù)Cosine similarity余弦相似度Cost curve成本曲線Cost Function成本函數(shù)Cost matrix成本矩陣Cost-sensitive成本敏感Cross entropy交叉嫡Cross validation交叉驗證Crowdsourcing眾包Curse of dimensionality維數(shù)

9、災難Cut point 截斷點Cutting plane algorithm割平面法Letter DData mining數(shù)據(jù)挖掘Data set數(shù)據(jù)集Decision Boundary決策邊界Decision stump決策樹樁Decision tree決策樹/判定樹Deduction 演繹深度卷積生成對抗網(wǎng)絡Deep Belief Network深度信念網(wǎng)絡Deep Convolutional Generative Adversarial Network/DCGANDeep learning深度學習Deep neural network/DNN深度神經(jīng)網(wǎng)絡Deep Q-Learning深度

10、 Q 學習Deep Q-Network深度 Q 網(wǎng)絡Density estimation密度估計Density-based clustering密度聚類Differentiable neural computer可微分神經(jīng)計算機Dimensionality reduction algorithm降維算法Directed edge有向邊Disagreement measure不合度量Discriminative model判另1J模型Discriminator判另1J器Distance measure品巨離度量Distance metric learning距離度量學習Distribution

11、分布Divergence 散度Diversity measure多樣性度量/差異性度量Domain adaption 領(lǐng)域自適應Downsampling 下采樣D-separation (Directed separation ) 有向分離Dual problem對偶問題Dummy node 啞結(jié)點Dynamic Fusion動態(tài)融合Dynamic programming 動態(tài)規(guī)戈 ULetter EEigenvalue decomposition特征值分解Embedding 嵌入Emotional analysis情緒分析Empirical conditional entropy經(jīng)驗條件嫡E

12、mpirical entropy經(jīng)驗嫡Empirical error經(jīng)驗誤差Empirical risk經(jīng)驗風險End-to-End 端到端Energy-based model基于能量的模型Ensemble learning集成學習Ensemble pruning集成修剪Error Correcting Output Codes/ ECOC糾錯輸出碼Error rate 錯誤率Error-ambiguity decomposition誤差-分歧分解Euclidean distance 歐氏品巨離Evolutionary computation演化計算Expectation-Maximizati

13、on期望最大化Expected loss期望損失Exploding Gradient Problem梯度爆炸問題Exponential loss function指數(shù)損失函數(shù)Extreme Learning Machine/ELM超限學習機Letter FFactorization因子分解False negative假負類False positive假正類False Positive Rate/FPR假正例率Feature engineering特征工程Feature selection特征選擇Feature vector特征向量Featured Learning特征學習Feedforward

14、 Neural Networks/FNN 前饋神經(jīng)網(wǎng)絡Fine-tuning 微調(diào)Flipping output 翻轉(zhuǎn)法Fluctuation 震蕩Forward stagewise algorithm 前向分步算法Frequentist 頻率主義學派Full-rank matrix滿秩矩陣Functional neuron功能神經(jīng)元Letter GGain ratio 增益率Game theory 博弈論Gaussian kernel function Gaussian Mixture Model General Problem Solving Generalization 泛化 Gener

15、alization error高斯核函數(shù)高斯混合模型通用問題求解泛化誤差Generalization error boundGeneralized Lagrange functionGeneralized linear modelGeneralized Rayleigh quotient泛化誤差上界廣義拉格朗日函數(shù) 廣義線性模型廣義瑞利商Generative Adversarial Networks/GAN生成對抗網(wǎng)絡Generative Model生成模型Generator 生成器Genetic Algorithm/GA遺傳算法Gibbs sampling 吉布斯采樣Gini index基尼

16、指數(shù)Global minimum全局最小Global Optimization全局優(yōu)化Gradient boosting梯度提升Gradient Descent梯度下降Graph theory圖論Ground-truth真相/真實Letter HHard margin 硬間隔Hard voting硬投票Harmonic mean調(diào)和平均Hesse matrix海塞矩陣Hidden dynamic model隱動態(tài)模型Hidden layer隱藏層Hidden Markov Model/HMM 隱馬爾可夫模型Hierarchical clustering層次聚類Hilbert space希爾伯特

17、空間Hinge loss function合頁損失函數(shù)Hold-out 留出法Homogeneous 同質(zhì)Hybrid computing混合計算Hyperparameter 超參數(shù)Hypothesis 假設Hypothesis test 假設驗證Letter IICML 國際機器學習會議Improved iterative scaling/IIS改進的迭代尺度法Incremental learning增量學習Independent and identically distributed/i.i.d.獨立同分布Independent Component Analysis/ICA 獨立成分分析I

18、ndicator function指示函數(shù)Individual learner個體學習器Induction 歸納Inductive bias歸納偏好Inductive learning歸納學習Inductive Logic Programming /ILP歸納邏輯程序設計Information entropy信息嫡Information gain信息增益Input layer 輸入層Insensitive loss不敏感損失Inter-cluster similarity簇間相似度International Conference for Machine Learning/ICML國際機器學習大

19、會Intra-cluster similarity簇內(nèi)相似度Intrinsic value 固有值Isometric Mapping/Isomap等度量映射Isotonic regression等分回歸Iterative Dichotomiser 迭代二分器Letter KKernel method核方法Kernel trick核技巧Kernelized Linear Discriminant Analysis/KLDA核線性判別分析K-fold cross validation k折交叉驗證/ k倍交叉驗證K-Means Clustering K -均值聚類K-Nearest Neighbo

20、urs Algorithm/KNNK 近鄰算法Knowledge base 知識庫Knowledge Representation 知識表征Letter LLabel space 標t己空間Lagrange duality拉格朗日對偶性Lagrange multiplier拉格朗日乘子Laplace smoothing拉普拉斯平滑Laplacian correction拉普拉斯修正Latent Dirichlet Allocation隱狄利克雷分布Latent semantic analysis潛在語義分析Latent variable隱變量Lazy learning懶惰學習Learner 學

21、習器Learning by analogy 類比學習Learning rate 學習率Learning Vector Quantization/LVQ學習向量量化Least squares regression tree最二乘回歸樹Leave-One-Out/LOO 留一法linear chain conditional random field線性鏈條件隨機場Linear Discriminant Analysis / LDA線性判別分析Linear model線性模型Linear Regression線性回歸Link function聯(lián)系函數(shù)Local Markov property局部馬

22、爾可夫性Local minimum局部最小Log likelihood對數(shù)似然Log odds /logit對數(shù)幾率Logistic Regression Logistic 回歸Log-likelihood對數(shù)似然Log-linear regression對數(shù)線性回歸Long-Short Term Memory/LSTM長短期記憶Loss function損失函數(shù)Letter MMachine translation/MT機器翻譯Macron-P宏查準率Macron-R宏查全率Majority voting 絕對多數(shù)投票法Manifold assumption流形假設Manifold lear

23、ning流形學習Margin theory間隔理論Marginal distribution邊際分布Marginal independence邊際獨立性Marginalization邊際化Markov Chain Monte Carlo/MCMC馬爾可夫鏈蒙特卡羅方法Markov Random Field馬爾可夫隨機場Word資料Maximal clique 最大團Maximum Likelihood Estimation/MLEMaximum margin最大間隔Maximum weighted spanning treeMax-Pooling 最大池化Mean squared error均

24、方誤差Meta-learner元學習器Metric learning度量學習Micro-P 微查準率Micro-R微查全率Minimal Description Length/MDLMinimax game 極小極大博弈極大似然估計/極大似然法最大帶權(quán)生成樹最小描述長度Misclassification cost誤分類成本Mixture of experts 混合專家Momentum 動量Moral graph道德圖/端正圖Multi-class classification多分類Multi-document summarization多文檔摘要Multi-layer feedforward

25、neural networks多層前饋神經(jīng)網(wǎng)絡Multilayer Perceptron/MLP多層感知器Multimodal learning多模態(tài)學習Multiple Dimensional Scaling多維縮放Multiple linear regression多元線性回歸Multi-response Linear Regression /MLR 多響應線性回歸Mutual information 互信息Letter NNaive bayes 樸素貝葉斯Naive Bayes Classifier樸素貝葉斯分類器Named entity recognition命名實體識別Nash eq

26、uilibrium納什均衡Natural language generation/NLG自然語言生成Natural language processing自然語言處理Negative class 負類Negative correlation負相關(guān)法Negative Log Likelihood負對數(shù)似然Neighbourhood Component Analysis/NCA近鄰成分分析Neural Machine Translation神經(jīng)機器翻譯Neural Turing Machine 神經(jīng)圖靈機Newton method 牛頓法NIPS 國際神經(jīng)信息處理系統(tǒng)會議No Free Lunch

27、 Theorem / NFL沒有免費的午餐定理Noise-contrastive estimation噪音對比估計Nominal attribute列名屬性Non-convex optimization 非凸優(yōu)化Nonlinear model非線性模型Non-metric distance非度量距離Non-negative matrix factorization非負矩陣分解Non-ordinal attribute無序?qū)傩訬on-Saturating Game非飽和博弈Norm 范數(shù)Normalization歸一化Nuclear norm核范數(shù)Numerical attribute數(shù)值屬性L

28、etter OObjective function目標函數(shù)Oblique decision tree斜決策樹Occam s razor奧卡姆剃刀Odds 幾率Off-Policy 離策略One shot learning一次性學習One-Dependent Estimator/ODE獨依賴估計On-Policy 在策略Ordinal attribute有序?qū)傩設ut-of-bag estimate包外估計Output layer輸出層Output smearing輸出調(diào)制法Overfitting過擬合/過配Oversampling過采樣Letter PPaired t-test成對 t 檢驗P

29、airwise成對型Pairwise Markov property成對馬爾可夫性Parameter參數(shù)Parameter estimation參數(shù)估計Parameter tuning 調(diào)參Parse tree 解析樹Particle Swarm Optimization/PSO粒子群優(yōu)化算法Part-of-speech tagging詞性標注Perceptron 感知機Performance measure性能度量Plug and Play Generative Network即插即用生成網(wǎng)絡Plurality voting相對多數(shù)投票法Polarity detection極性檢測Polyn

30、omial kernel function多項式核函數(shù)Pooling 池化Positive class正類Positive definite matrix正定矩陣Post-hoc test后續(xù)檢驗Post-pruning后剪枝potential function 勢函數(shù)Precision查準率/準確率Prepruning預剪枝Principal component analysis/PCA主成分分析Principle of multiple explanations多釋原貝 UPrior 先驗Probability Graphical Model概率圖模型Proximal Gradient D

31、escent/PGD近端梯度下降Pruning 剪枝Pseudo-label 偽標記Letter QQuantized Neural Network量子化神經(jīng)網(wǎng)絡Quantum computer量子計算機Quantum Computing量子計算Quasi Newton method擬牛頓法Letter RRadial Basis Function / RBF 徑向基函數(shù)Random Forest Algorithm隨機森林算法Random walk隨機漫步受試者工作特征Recall查全率/召回率Receiver Operating Characteristic/ROCRectified Li

32、near Unit/ReLU線性修正單元Recurrent Neural Network循環(huán)神經(jīng)網(wǎng)絡Recursive neural network遞歸神經(jīng)網(wǎng)絡Reference model參考模型Regression 回歸Regularization正貝 U 化Reinforcement learning/RL強化學習Representation learning表征學習Representer theorem表示定理reproducing kernel Hilbert space/RKHS再生核希爾伯特空間Re-sampling 重采樣法Rescaling 再縮放Residual Mappi

33、ng殘差映射受限玻爾茲曼機限定等距性Residual Network殘差網(wǎng)絡Restricted Boltzmann Machine/RBMRestricted Isometry Property/RIPRe-weighting重賦權(quán)法Robustness 穩(wěn)健性/魯棒性Root node 根結(jié)點Rule Engine 規(guī)則引擎Rule learning規(guī)則學習Letter SSaddle point鞍點Sample space樣本空間Sampling 采樣Score function評分函數(shù)Self-Driving自動駕駛Self-Organizing Map /SOM自組織映射Semi-n

34、aive Bayes classifiers半樸素貝葉斯分類器Semi-Supervised Learning半監(jiān)督學習semi-Supervised Support Vector Machine半監(jiān)督支持向量機Sentiment analysis情感分析Separating hyperplane 分離超平面Sigmoid functionSigmoid 函數(shù)Similarity measure相似度度量Simulated annealing模擬退火Simultaneous localization and mapping同步定位與地圖構(gòu)建Singular Value Decompositio

35、n奇異值分解Slack variables松弛變量Smoothing 平滑Soft margin 軟間隔Soft margin maximization軟間隔最大化Soft voting軟投票Sparse representation 稀疏表征Sparsity稀疏性Specialization 特化Spectral Clustering譜聚類Speech Recognition語音識另1JSplitting variable切分變量Squashing function擠壓函數(shù)Stability-plasticity dilemma可塑性-穩(wěn)定性困境Statistical learning統(tǒng)計學習Status feature function狀態(tài)特征函Stochastic gradient descent 隨機梯度下降Stratified sampling分層采樣Structural risk結(jié)構(gòu)風險Structural risk minimization/SRM結(jié)構(gòu)風險最小化Subspace 子空間Supervised learning監(jiān)督學習/有導師學習support vector expansion支持向量展式Support Vector Mach