Definition
Double descent in statistics and machine learning is the phenomenon where a model's error rate on the test set initially decreases with the number of parameters, then peaks, then decreases again. This phenomenon has been considered surprising, as it contradicts assumptions about overfitting in classical machine learning.
Related concepts
15.aiAAAI Conference on Artificial IntelligenceAI agentAI alignmentAI anthropomorphismAI boomAI bubbleAI data centerAI effectAI literacyAI nationalismAI safetyAI slopAI takeoverAI veganismAI winterAccelerated failure time modelAction selectionActivation functionActive learning (machine learning)Actuarial scienceAdaptive clinical trialAdobe FireflyAdversarial machine learningAgent2AgentAidan GomezAkaike information criterionAlan TuringAlexNetAlex Graves (computer scientist)Alex KrizhevskyAlgorithmic probabilityAllen NewellAlphaFoldAlphaGoAlphaZeroAnalysis of covarianceAnalysis of varianceAnderson–Darling testAndrej KarpathyAndrew NgAnomaly detectionAnscombe transformApplications of artificial intelligenceApprenticeship learningArithmetic meanArithmetic–geometric meanArtificial Intelligence ActArtificial Intelligence Cold WarArtificial general intelligenceArtificial human companionArtificial intelligenceArtificial intelligence and electionsArtificial intelligence arms raceArtificial intelligence in architectureArtificial intelligence in educationArtificial intelligence in fictionArtificial intelligence in healthcareArtificial intelligence in mental healthArtificial intelligence in video gamesArtificial intelligence visual artArtificial neural networkArtificial superintelligenceAshish VaswaniAssociation rule learningAsymptotic theory (statistics)Attention (machine learning)Aurora (text-to-image model)AutoGPTAutocorrelationAutoencoderAutomated machine learningAutomated reasoningAutomated theorem provingAutoregressive conditional heteroskedasticityAutoregressive modelAutoregressive–moving-average modelAverage absolute deviationBERT (language model)BIRCHBLOOM (language model)BackpropagationBar chartBatch learningBatch normalizationBayes estimatorBayes factorBayesian inferenceBayesian information criterionBayesian linear regressionBayesian networkBayesian probabilityBernard WidrowBias of an estimatorBias–variance tradeoffBinomial regressionBioinformaticsBiostatisticsBiplotBlocking (statistics)Boltzmann machineBoosting (machine learning)Bootstrap aggregatingBootstrapping (statistics)Box plotBox–Cox transformationBox–Jenkins methodBreusch–Godfrey testCURE algorithmCanonical correlationCartographyCategorical variableCensusCentral limit theoremCentral tendencyChatbot psychosisChemometricsChi-squared testChinchilla (language model)Christopher D. ManningClaude (language model)Claude ShannonCliff ShawClinical study designClinical trialCluster analysisCluster samplingCochran–Mantel–Haenszel statisticsCoefficient of determinationCoefficient of variationCohen's kappaCohort studyCointegrationCommunications on Pure and Applied MathematicsCompetition in artificial intelligenceCompleteness (statistics)Computational learning theoryComputer visionConditional random fieldConference on Neural Information Processing SystemsConfidence intervalConfoundingConfusion matrixConjugate gradient methodContingency tableContinuous probability distributionContraharmonic meanControl chartConvolutionConvolutional neural networkCorrelation and dependenceCorrelogramCount dataCredible intervalCrime statisticsCross-correlationCross-sectional studyCross-validation (statistics)CrowdsourcingCubic meanCurriculum learningDALL-EDBRXDBSCANDaniel Kokotajlo (researcher)Data augmentationData cleaningData collectionData miningData preprocessingData transformation (statistics)David Silver (computer scientist)Decision tree learningDecomposition of time seriesDeepDreamDeepSeek (chatbot)Deep learningDeep learning speech synthesisDegrees of freedom (statistics)Demis HassabisDemographic statisticsDensity estimationDescriptive statisticsDesign of experimentsDetrendingDickey–Fuller testDifferencingDifferentiable neural computerDiffusion modelDiffusion processDimensionality reductionDivergence (statistics)Dream Machine (text-to-video model)Durbin–Watson statisticECML PKDDEcho state networkEconometricsEffect sizeEffective dimensionEfficiency (statistics)Electrochemical RAMElevenLabsElliptical distributionEmpirical distribution functionEmpirical risk minimizationEngineering statisticsEnsemble learningEnvironmental impact of artificial intelligenceEnvironmental statisticsEpidemiologyErrors and residualsEstimating equationsEthics of artificial intelligenceExpectation–maximization algorithmExperimentExplainable artificial intelligenceExponential familyExponential smoothingF-testFacial recognition systemFactor analysisFactorial experimentFailure rateFan chart (statistics)Feature engineeringFeature learningFeature scalingFeedforward neural networkFei-Fei LiFirst-hitting-time modelFisher transformationFlux (text-to-image model)Forest plotFourier analysisFrank RosenblattFrançois CholletFrequency distributionFrequency domainFrequentist inferenceFriedman testFuzzy clusteringG-testGPT ImageGated recurrent unitGating mechanismGaussian processGemini (chatbot)Gemini (language model)Gemma (language model)General linear modelGeneralized linear modelGeneralized meanGenerative AIGenerative adversarial networkGenerative engine optimizationGenerative modelGenerative pre-trained transformerGenie (world model)Geoffrey HintonGeographic information systemGeometric meanGeostatisticsGloVeGlossary of artificial intelligenceGoodness of fitGradient descentGrammar inductionGranger causalityGraph neural networkGraphical modelGrok (chatbot)Grokking (machine learning)Grouped dataHallucination (artificial intelligence)Handwriting recognitionHarmonic meanHeatmapHeinz meanHerbert A. SimonHeronian meanHessian matrixHidden Markov modelHierarchical clusteringHighway networkHistogramHistory of artificial intelligenceHodges–Lehmann estimatorHomoscedasticity and heteroscedasticityHuawei PanGuHuman-in-the-loopHuman image synthesisHumanity's Last ExamHyperparameter (machine learning)IBM GraniteIBM WatsonIBM WatsonxIan GoodfellowIdeogram (text-to-image model)Ilya SutskeverImagen (text-to-image model)Imitation learningIndependent component analysisIndex of dispersionIntelligent agentInteraction (statistics)International Conference on Learning RepresentationsInternational Conference on Machine LearningInternational Joint Conference on Artificial IntelligenceInterquartile rangeInterval estimationIsolation forestIsotonic regressionIsotropyJackknife resamplingJames GoodnightJan LeikeJarque–Bera testJohansen testJohn HopfieldJohn McCarthy (computer scientist)John SchulmanJohn von NeumannJonckheere's trend testJoseph WeizenbaumJournal of Machine Learning ResearchJurimetricsJürgen SchmidhuberK-means clusteringK-nearest neighbors algorithmKaplan–Meier estimatorKendall rank correlation coefficientKernel machinesKinshipKling AIKolmogorov–Smirnov testKrigingKruskal–Wallis testKunihiko FukushimaKurtosisL-momentLaMDALanguage modelLarge language modelLatent diffusion modelLeNetLearning curve (machine learning)Learning to rankLeast-squares spectral analysisLehmann–Scheffé theoremLehmer meanLethal autonomous weaponLikelihood-ratio testLikelihood functionLikelihood intervalLilliefors testLine chartLinear discriminant analysisLinear regressionList of artificial intelligence companiesList of artificial intelligence projectsList of datasets for machine-learning researchList of datasets in computer vision and image processingList of fields of application of statisticsList of statistical testsList of statistics articlesLjung–Box testLlama (language model)Local outlier factorLocation parameterLocation–scale familyLog-rank testLog transformationLogistic regressionLong short-term memoryLoss functionLoss functions for classificationLotfi A. ZadehLp spaceM-estimatorMachine Learning (journal)Machine learningMamba (deep learning architecture)Mann–Whitney U testMarvin MinskyMaximum a posteriori estimationMaximum likelihoodMcNemar's testMeanMean shiftMechanistic interpretabilityMedianMedian-unbiased estimatorMedical statisticsMemtransistorMeta-learning (computer science)Method of moments (statistics)Methods engineeringMidjourneyMiniMax (company)Minimum-variance unbiased estimatorMinimum distance estimationMin–max normalizationMissing dataMixed modelMode (statistics)Model Context ProtocolModel selectionModel specificationMoment (mathematics)Monotone likelihood ratioMuZeroMulti-agent reinforcement learningMultilayer perceptronMultimodal learningMultiple comparisonsMultivariate adaptive regression splinesMultivariate analysis of varianceMultivariate distributionMultivariate normal distributionMultivariate statisticsMusic and artificial intelligenceMustafa SuleymanNaive Bayes classifierNathaniel Rochester (computer scientist)National accountsNatural experimentNelson–Aalen estimatorNeural Turing machineNeural fieldNeural machine translationNeural network (machine learning)Neural radiance fieldNeuro-symbolic AINeuromorphic engineeringNoam ShazeerNon-negative matrix factorizationNonlinear regressionNonparametric regressionNonparametric statisticsNormalization (machine learning)Normalization (statistics)OPTICS algorithmOasis (Minecraft clone)Observational studyOccam learningOfficial statisticsOliver SelfridgeOne- and two-tailed testsOnline machine learningOntology learningOpenAI FiveOpinion pollOptical character recognitionOptimal decisionOptimal designOrder statisticOrdinary least squaresOriol VinyalsOutlierOutline of machine learningOutline of statisticsOverfittingPaLMParameterParametric statisticsPartial autocorrelation functionPartial correlationPartition of sums of squaresPaul WerbosPearson correlation coefficientPearson product-moment correlation coefficientPercentilePerceptronPermutation testPhysics-informed neural networksPie chartPivotal quantityPlug-in principlePoint estimationPoisson regressionPolicy gradient methodPopulation (statistics)Population statisticsPosterior probabilityPower (statistics)Power transformPrecautionary principlePrediction intervalPrincipal component analysisPrior probabilityProbabilistic designProbability distributionProbably approximately correct learningProject DebaterPrompt engineeringProper generalized decompositionProportional hazards modelPsychometricsQ-learningQuality controlQuantum machine learningQuasi-Newton methodQuasi-experimentQuestionnaireQuoc V. LeQwenQ–Q plotRadar chartRandom assignmentRandom forestRandom sample consensusRandomization testRandomized controlled trialRandomized experimentRange (statistics)Rank correlationRanking (statistics)Rao–Blackwell theoremReasoning modelReceiver operating characteristicRecraftRectifier (neural networks)Recurrent neural networkRecursive self-improvementReflection (artificial intelligence)Regression analysisRegression validationRegularization (mathematics)Regulation of artificial intelligenceRegulation of artificial intelligence in the United StatesReinforcement learningReinforcement learning from human feedbackRelevance vector machineReliability engineeringReplica trickReplication (statistics)Resampling (statistics)Reservoir computingResidual neural networkRestricted Boltzmann machineRetrieval-augmented generationRiffusionRobot controlRobust regressionRobust statisticsRule-based machine learningRun chartRunway (company)SIAM Journal on Mathematics of Data ScienceSample medianSample size determinationSampling (statistics)Sampling distributionScale parameterScatter plotScientific controlScore testSeasonal adjustmentSeedance 2.0Self-driving carSelf-organizing mapSelf-play (reinforcement learning technique)Self-supervised learningSemantic analysis (machine learning)Semi-supervised learningSemiparametric regressionSeppo LinnainmaaSeq2seqSeymour PapertShape of the distributionShape parameterShapiro–Wilk testShun'ichi AmariSigmoid functionSign testSimple linear regressionSimultaneous equations modelSkewnessSocial statisticsSoftmax functionSora (text-to-video model)Sparse dictionary learningSpatial analysisSpearman's rank correlation coefficientSpectral density estimationSpeech recognitionSpiking neural networkStable DiffusionStandard deviationStandard errorStandard scoreState–action–reward–state–actionStationary processStatisticStatistical classificationStatistical dispersionStatistical distanceStatistical graphicsStatistical hypothesis testStatistical inferenceStatistical learning theoryStatistical modelStatistical parameterStatistical populationStatistical powerStatistical process controlStatistical theoryStatisticsStem-and-leaf displayStephen GrossbergStochastic approximationStochastic gradient descentStratified samplingStructural breakStructural equation modelingStructured predictionStuart GemanStudent's t-testSufficient statisticSuno (platform)Supervised learningSupport vector machineSurvey methodologySurvival analysisSurvival functionSymbolic artificial intelligenceSystem identificationT-distributed stochastic neighbor embeddingT5 (language model)Takeo KanadeTemporal difference learningTest setText-to-image modelText-to-video modelThermodynamic limitTime domainTime seriesTimeline of artificial intelligenceTolerance intervalTopological deep learningTraining, validation, and test data setsTransformer (deep learning)Transformer (deep learning architecture)Trend estimationTruncation (statistics)U-NetU-statisticUdioUncanny valleyUniformly most powerful testUnit vector normalizationUnsupervised learningUp-and-Down DesignsV-statisticVan der Waerden testVapnik–Chervonenkis theoryVarianceVariance-stabilizing transformationVariational autoencoderVector autoregressionVeo (text-to-video model)Vibe codingViolin plotVirtual politicianVision transformerWald testWalter PittsWarren Sturgis McCullochWaveNetWaveletWeak artificial intelligenceWeight initializationWhisper (speech recognition system)Whittle likelihoodWilcoxon signed-rank testWinsorizingWord2vecWord embeddingWorkplace impact of artificial intelligenceWorld model (artificial intelligence)Xiaomi MiMoYann LeCunYeo–Johnson transformationYoshua BengioZ-test
27 concepts already in your glossary