Statistics

Causal Inference Under Hidden Confounding

Development of methods to identify and estimate causal effects when unmeasured confounders may bias traditional causal inference approaches.

RESEARCH GAP FRONTIERS

High-Dimensional Variable Selection

Advanced techniques for identifying important predictors in settings where the number of variables far exceeds sample size.

RESEARCH GAP FRONTIERS

Functional Data Analysis Methods

Statistical approaches for analyzing data that are curves, surfaces, or other functional objects rather than scalar measurements.

RESEARCH GAP FRONTIERS

Bayesian Nonparametric Models

Development of Bayesian inference methods that avoid specifying finite-dimensional parameter spaces using techniques like Dirichlet processes.

RESEARCH GAP FRONTIERS

Network Analysis and Graph Statistics

Statistical methods for analyzing complex networks and graph-structured data with applications to social and biological networks.

RESEARCH GAP FRONTIERS

Robust Statistics and Outlier Detection

Development of statistical procedures resistant to deviations from model assumptions and contaminated data.

RESEARCH GAP FRONTIERS

Time Series Forecasting Deep Learning

Integration of neural network architectures with time series statistical theory for improved long-horizon predictions.

RESEARCH GAP FRONTIERS

Spatial Statistics and Geostatistics

Methods for modeling and predicting spatially correlated data using kriging, point processes, and spatial regression techniques.

RESEARCH GAP FRONTIERS

Survival Analysis with Competing Risks

Advanced techniques for analyzing time-to-event data when subjects face multiple mutually exclusive failure modes.

Machine Learning Theory and Generalization

Theoretical analysis of statistical learning theory, sample complexity, and generalization bounds for machine learning algorithms.

Differential Privacy and Data Protection

Statistical methods that provide formal privacy guarantees while enabling accurate analysis of sensitive data.

Causal Discovery from Observational Data

Algorithms for discovering causal directed acyclic graphs from observational data without experimental intervention.

Semiparametric Efficiency and Influence Functions

Theory and methods for achieving efficiency bounds in semiparametric models using doubly robust estimation.

Quantile Regression Advances

Extension of quantile regression methods to high-dimensional settings and nonlinear models for complete distributional analysis.

Reinforcement Learning and Bandits

Statistical methods for sequential decision-making under uncertainty including multi-armed bandits and contextual bandits.

Missing Data and Imputation Methods

Development of principled approaches to handle incomplete data through multiple imputation and modern missing mechanisms.

Graphical Models and Markov Networks

Statistical methods for learning and inference in probabilistic graphical models including Bayesian and Markov networks.

Adaptive Experimental Design

Development of sequential and adaptive designs that modify future treatments based on accumulating data.

Optimal Transport and Wasserstein Methods

Application of optimal transport theory to statistical problems including distribution matching and generative modeling.

Reproduceability and Statistical Power

Methods for improving research reproducibility through better experimental design, power analysis, and multiple testing correction.

Compositional Data Analysis

Statistical techniques for analyzing data constrained to sum to a constant, common in microbiome and genomic studies.

Preferential Attachment and Stochastic Blockmodels

Statistical models for network formation and community detection in complex networks.

Dimension Reduction via Manifold Learning

Methods for discovering and projecting data onto lower-dimensional manifolds while preserving intrinsic geometric structure.

Scalable Bayesian Computation and Variational Inference

Development of computationally efficient approximations to Bayesian posteriors for large datasets.

Conformal Prediction and Uncertainty Quantification

Distribution-free methods for constructing prediction intervals with guaranteed coverage properties.

Statistical Learning on Graphs

Theory and methods for supervised and unsupervised learning on graph-structured data and node features.

Mixture Models and Clustering

Statistical methods for learning mixture models and clustering including EM algorithms and model selection.

Longitudinal Data Analysis

Methods for analyzing repeated measures and longitudinal data with complex correlation structures.

Extreme Value Statistics and Tail Risk

Theory and methods for analyzing rare extreme events and tail distributions in univariate and multivariate settings.

Count Data and Poisson Models

Statistical models for count outcomes including zero-inflation, overdispersion, and negative binomial extensions.

Copula Methods and Dependence

Statistical techniques using copulas to model dependence structures between variables independently of margins.

Online Learning and Streaming Data

Algorithms for learning from sequentially arriving data with limited memory and computational resources.

Empirical Bayes and Hierarchical Modeling

Methods for learning hyperparameters from data in hierarchical Bayesian models and empirical Bayes estimation.

Transfer Learning and Domain Adaptation

Statistical methods for leveraging knowledge from source domains to improve prediction in target domains.

Point Process Theory and Hawkes Processes

Analysis of self-exciting point processes and modeling of arrival times with clustering behavior.

Personalized Medicine and Precision Treatment

Statistical methods for identifying optimal treatment rules tailored to individual patient characteristics.

Debiased Machine Learning Estimators

Methods for obtaining valid statistical inference despite using machine learning for nuisance parameter estimation.

Multiple Testing and False Discovery Rate

Procedures for controlling error rates in multiple hypothesis testing including FDR and adaptive methods.

Truncated and Censored Data

Statistical inference methods for data with truncation or censoring mechanisms common in survival analysis.

Heterogeneous Treatment Effect Estimation

Methods for identifying and estimating how treatment effects vary across subpopulations and individuals.

Topological Data Analysis

Statistical methods using topological and geometric tools including persistent homology for data analysis.

Neural Network Generalization Theory

Theoretical analysis of deep learning generalization, overparameterization, and implicit regularization.

Covariate Balance and Propensity Scores

Methods for achieving balance in confounding variables using propensity scores and weighting techniques.

High-Dimensional Covariance Estimation

Regularized methods for estimating covariance matrices when dimensions exceed sample size.

Bayesian Optimization and Acquisition Functions

Statistical methods for sequential optimization of expensive black-box functions using surrogate models.

Knockoffs and Feature Importance

Methods for variable importance and feature selection using knockoff variables with false discovery control.

Generalized Additive Models and Splines

Flexible semiparametric regression using basis functions and smoothing splines for nonlinear relationships.

Bayesian Nonparametric Density Estimation

Development of flexible Bayesian methods for estimating unknown probability densities without parametric assumptions.

Data Integration and Fusion Methods

Statistical techniques for combining information from multiple heterogeneous data sources and modalities.

Survival Trees and Random Forests

Tree-based ensemble methods adapted for time-to-event analysis and censored data.

Causal Forests and Heterogeneous Effects

Development and analysis of random forest methods for estimating individualized causal effects across subgroups in observational and experimental data.

Sequential Hypothesis Testing and Optimal Stopping

Theory and methods for conducting statistical tests with flexible sample sizes and sequential decision-making under uncertainty.

Matrix Completion and Low-Rank Recovery

Statistical theory for reconstructing high-dimensional matrices from incomplete observations using convex optimization and rank constraints.

Causal Mediation Analysis and Indirect Effects

Methods for decomposing causal effects into direct and indirect pathways through intermediate variables in complex systems.

Anomaly Detection in High-Dimensional Data

Statistical approaches for identifying unusual patterns and outliers in massive high-dimensional datasets using projection and clustering techniques.

Subgroup Analysis and Interaction Effects

Methods for identifying and validating differential treatment effects across patient subgroups in clinical trials and observational studies.

Sparsity and Regularization in Linear Models

Theory and algorithms for LASSO, elastic net, and related penalized regression methods for variable selection in high dimensions.

Instrumental Variables and Two-Stage Least Squares

Statistical methods for causal estimation when endogeneity is present and valid instrumental variables are available.

Kernel Methods and Reproducing Kernel Hilbert Spaces

Theory of kernels and RKHS methods for nonlinear statistical learning including support vector machines and kernel ridge regression.

Isotonic Regression and Order-Constrained Estimation

Statistical methods for estimating monotone relationships and functions subject to order constraints in data.

U-Statistics and Von Mises Expansions

Theory of U-statistics for robust estimation with applications to nonparametric testing and asymptotic distribution theory.

Stochastic Approximation and Recursive Algorithms

Convergence analysis of iterative algorithms for optimization and learning in online and streaming data settings.

Regression Discontinuity Design and Local Methods

Causal inference methods exploiting sharp or fuzzy threshold discontinuities using local polynomial and bandwidth selection techniques.

Empirical Process Theory and Symmetry Arguments

Mathematical foundations for understanding convergence of empirical distributions with applications to machine learning generalization.

Network Intervention and Graph Experiments

Theory for designing and analyzing randomized experiments on networks with spillover effects and interference.

Nonparametric Goodness-of-Fit Testing

Development of tests for distributional assumptions using empirical processes, permutation methods, and kernel approaches.

Causal Assumptions and Identifiability Verification

Methods for validating and testing the plausibility of causal identification assumptions in observational research.

Boosting Algorithms and Ensemble Methods

Theory and practice of iterative reweighting and ensemble learning methods including AdaBoost and gradient boosting.

Wavelet Analysis and Time-Frequency Methods

Statistical methods using wavelets for non-stationary signal analysis, denoising, and feature extraction.

Double Machine Learning and Orthogonal Estimation

Framework for combining machine learning with semiparametric efficiency to estimate low-dimensional functionals robustly.

Change Point Detection and Structural Breaks

Methods for identifying and estimating locations where the statistical properties of time series or sequential data change.

Inverse Problems and Deconvolution

Statistical techniques for recovering unobservable quantities from indirect or degraded measurements with regularization.

Causal Inference with Time-Varying Treatments

Methods for estimating causal effects when treatments are assigned sequentially over time with feedback loops.

Nonparametric Hypothesis Testing and Rank Tests

Distribution-free testing procedures based on ranks and order statistics for comparing samples without distributional assumptions.

Information-Theoretic Bounds and Limits

Fundamental limits on statistical estimation and learning using entropy, mutual information, and minimax theory.

Stratification and Conditional Inference

Methods for conducting statistical analysis within strata defined by covariates to improve efficiency and validity.

High-Dimensional Testing and Multiple Comparisons

Procedures for simultaneously testing many hypotheses in high dimensions while controlling family-wise error rate.

Causal Models and Structural Equation Modeling

Framework for specifying directed acyclic graphs and latent variables to represent causal relationships in systems.

Resampling Methods and Bootstrap Theory

Theory and applications of bootstrap, jackknife, and permutation methods for inference without distributional assumptions.

Model Selection and Cross-Validation

Methods for selecting among competing models using sample splitting, leave-one-out validation, and generalization bounds.

Bayesian Variable Selection and Model Averaging

Bayesian approaches for uncertainty in variable selection with marginal likelihood computation and posterior model probabilities.

Density Ratio Estimation and Importance Weighting

Methods for estimating ratios of probability densities with applications to covariate shift and distribution matching.

Functional Time Series and Temporal Dynamics

Statistical analysis of curves and functions observed sequentially over time with temporal dependence.

Monotone Likelihood Ratio and Stochastic Ordering

Theory of likelihood orderings for constructing optimal tests and inference with shape constraints.

Sieve Estimation and Smoothing Parameter Selection

Nonparametric estimation using basis expansions with automatic selection of smoothing and complexity parameters.

Approximate Bayesian Computation and Likelihood-Free Inference

Methods for Bayesian inference when likelihood is intractable using simulation-based acceptance-rejection algorithms.

Causal Graphs and D-Separation Criterion

Graphical methods for determining identifiability of causal effects using do-calculus and backdoor adjustments.

Shape-Constrained Estimation and Monotonicity

Estimating functions subject to convexity, concavity, or monotonicity constraints using projections and optimization.

Partial Linear and Index Models

Semiparametric models combining parametric and nonparametric components with dimension reduction techniques.

Semi-Supervised Learning and Self-Training

Statistical methods for learning from labeled and unlabeled data with theoretical guarantees under manifold assumptions.

Matching Estimators and Nearest Neighbor Methods

Nonparametric estimation using matching strategies for causal inference and local averaging with bandwidth theory.

Bayesian Credible Intervals and Frequentist Coverage

Construction of Bayesian intervals with guaranteed frequentist coverage properties for valid uncertainty quantification.

Recurrent Neural Networks and Sequence Models

Statistical theory for LSTM and GRU networks applied to sequential data with analysis of long-range dependencies.

Symmetry and Invariance in Statistical Models

Exploitation of symmetry groups and invariance principles for constructing optimal and equivariant estimators.

Sensitivity Analysis and Robustness Testing

Methods for assessing how inferences change under violations of key assumptions in causal and observational studies.

Accelerated Failure Time and Additive Hazards

Parametric and semiparametric survival models with alternative formulations to proportional hazards.

Nonparametric Bayes and Dirichlet Processes

Bayesian nonparametric models for mixture distributions and flexible functional estimation with posterior computation.

Attention Mechanisms and Transformer Networks

Statistical properties and learning theory for attention-based architectures in sequence-to-sequence modeling.

Conservative Confidence Intervals and Median Unbiasedness

Construction of intervals guaranteed to contain true parameters with minimal coverage assumptions and bias properties.

Statistical Mediation Analysis and Indirect Effects

Methods for decomposing total effects into direct and indirect pathways through mediator variables in causal systems.

Instrumental Variables and Mendelian Randomization

Techniques using genetic or natural instruments to identify causal effects when treatment assignment is not randomly assigned.

Targeted Maximum Likelihood Estimation

Doubly robust estimation framework combining machine learning with semiparametric theory for efficient causal parameter estimation.

Synthetic Controls and Panel Data Methods

Statistical approaches for causal inference in observational studies using weighted combinations of control units and longitudinal measurements.

Regression Discontinuity Design Theory

Theoretical foundations and extensions of quasi-experimental designs exploiting threshold effects in treatment assignment.

Difference-in-Differences with Multiple Treatments

Methodological advances for causal inference when multiple units receive treatments at different times in panel settings.

Distributed Statistical Computing and Federated Learning

Algorithms for parameter estimation and inference across decentralized data sources without centralized data aggregation.

Tensor Methods and Higher-Order Data

Statistical theory and algorithms for decomposing and analyzing multidimensional arrays beyond traditional matrix methods.

Selective Inference and Post-Selection Inference

Statistical methods accounting for selection bias when inference follows data-dependent model or variable selection procedures.

Causal Sensitivity Analysis and Bounds

Quantifying robustness of causal conclusions to violations of standard assumptions through partial identification and bounds.

Graphical Model Structure Learning

Methods for estimating sparse precision matrices and conditional independence graphs from high-dimensional data.

Causal Inference with Interference

Statistical frameworks addressing violations of stable unit treatment value assumption when treatments affect other units.

Mixture Models for Longitudinal Trajectories

Latent class models identifying heterogeneous developmental patterns and subgroup memberships in repeated measurements.

Sequential Testing and Group Sequential Designs

Statistical procedures allowing interim analysis and early stopping in clinical trials maintaining type I error control.

Deconvolution and Measurement Error Models

Methods for estimation when covariates or outcomes are observed with additive or multiplicative measurement noise.

Shape-Constrained Nonparametric Regression

Statistical estimation under constraints like monotonicity, convexity, or other shape restrictions on regression functions.

Causal Inference from Text and Natural Language

Methods for extracting causal relationships and estimating treatment effects from unstructured textual data and documents.

Sparse Principal Component Analysis

Dimension reduction techniques producing interpretable components with sparse loadings in high-dimensional settings.

Empirical Process Theory and U-Statistics

Foundational theory for consistency and asymptotic distribution of statistics based on empirical measures.

Matrix Completion and Missing Data in Matrices

Recovery of low-rank matrices from incomplete observations using convex optimization and statistical guarantees.

Dependent and Weak Dependent Data Analysis

Statistical methods for data with temporal or spatial dependence using mixing conditions and covariance structures.

Cross-Validation and Model Selection Theory

Theoretical foundations of resampling methods for hyperparameter selection and generalization error estimation.

Frailty Models and Random Effects Survival

Survival analysis extensions incorporating unobserved heterogeneity through random effects or frailty distributions.

Multivariate Functional Data Methods

Statistical techniques for analyzing multiple curves or surfaces simultaneously in functional data contexts.

Confidence Sets and Inversion of Tests

Constructing confidence regions through inverting statistical tests with optimal properties and coverage guarantees.

Spike and Slab Priors and Sparse Bayesian Methods

Bayesian variable selection using mixture priors encouraging sparsity in high-dimensional parameter estimation.

Stochastic Approximation and Gradient Descent

Convergence theory and optimization algorithms for finding solutions to implicit equations in online settings.

Causal Discovery with Latent Variables

Methods for learning directed acyclic graphs when some variables are unobserved but structure is identifiable.

Biomarker Discovery and Classification Trees

Statistical methods for identifying predictive biomarkers and constructing interpretable decision trees in precision medicine.

Isotonic Regression and Order Restrictions

Estimation under inequality constraints on parameters such as monotonicity or stochastic ordering assumptions.

Functional Time Series and Curve Registration

Methods for analyzing sequences of curves including alignment and forecasting of functional observations.

Graphical Lasso and Sparse Covariance

Penalized likelihood approaches for estimating sparse inverse covariance matrices and conditional independence structures.

Bootstrap Methods and Resampling Theory

Theoretical properties and applications of resampling procedures for estimation and inference without parametric assumptions.

Minimax Theory and Information Limits

Fundamental lower bounds and optimal rates for statistical estimation connecting information theory and statistics.

Generalized Linear Mixed Models Extensions

Advances in fitting and inference for regression models with both fixed and random effects in non-normal data.

Approximation Theory and Statistical Bases

Using basis expansions, wavelets, and splines for flexible nonparametric function estimation in statistical models.

Causal Inference in Network Experiments

Methods for estimating treatment effects in experiments where subjects are connected in networks with interference.

Specification Testing and Model Diagnostics

Goodness-of-fit tests and diagnostic tools for checking model assumptions and detecting misspecification.

Large-Scale Multiple Testing Procedures

Methods for controlling error rates in thousands or millions of simultaneous hypothesis tests.

Dimension Reduction for Regression Prediction

Techniques like sufficient dimension reduction extracting lower-dimensional projections predictive of response variables.

Bayesian Additive Regression Trees

Ensemble methods combining Bayesian inference with tree-based additive models for flexible nonparametric regression.

Influence Functions and Robustness Assessment

Computing and applying influence functions to understand estimator stability and quantify outlier impact.

Shrinkage Estimation and James-Stein Methods

Biased estimators with reduced mean squared error through shrinkage toward common targets or regularization.

Spectral Methods for Data Analysis

Eigenvalue and eigenvector-based techniques for clustering, community detection, and low-rank approximations.

Moment-Based Inference and Generalized Method of Moments

Estimating models defined by moment conditions without full likelihood specification using GMM and related approaches.

Extremal Inference and Multivariate Extremes

Statistical modeling of joint tail behavior and dependence among multiple variables in extreme regions.

Statistics of Shape and Landmark Data

Methods for analyzing configurations of points and shape variations in morphometric and geometric data.

Semi-Supervised Learning Theory and Applications

Theoretical foundations and practical algorithms for learning from partially labeled datasets in complex statistical settings.

Graphical Lasso and Sparse Precision Estimation

Advanced sparse estimation techniques for high-dimensional inverse covariance matrices with applications to network inference.

Change Point Detection in High Dimensions

Statistical methods for identifying structural breaks and regime changes in high-dimensional time series and sequential data.

Fairness and Algorithmic Bias in Statistics

Development of statistical frameworks ensuring equity and reducing discrimination in machine learning algorithms and predictions.

Multitask Learning and Shared Representations

Statistical theory for learning multiple related tasks simultaneously through shared feature representations and parameter structures.

Nonparametric Hypothesis Testing Methods

Development of distribution-free statistical tests for complex alternatives without parametric model assumptions.

Tensor Analysis and Multilinear Algebra

Statistical methods for decomposing and analyzing high-order tensor data with applications to neuroimaging and genomics.

Active Learning and Query Strategies

Optimal design of data acquisition strategies to maximize information gain with minimal labeling costs.

Multilevel Modeling and Hierarchical Structures

Statistical frameworks for analyzing nested data structures with variation at multiple levels of organization.

Zero-Inflated and Hurdle Models

Statistical methods for modeling count data with excess zeros common in ecological and biomedical applications.

Approximate Bayesian Computation Methods

Simulation-based inference techniques for complex models where likelihood functions are intractable or unavailable.

Subgroup Identification and Interaction Detection

Methods for discovering patient subgroups with differential treatment responses and detecting treatment-covariate interactions.

Empirical Likelihood and Generalized Estimating Equations

Semiparametric methods for inference in dependent data and complex survey sampling designs.

Spike-and-Slab Priors and Bayesian Selection

Bayesian variable selection methods using mixture priors for simultaneous model selection and estimation.

Recurrent Neural Networks and Sequence Modeling

Statistical analysis and theory of recurrent neural network architectures for temporal and sequential dependencies.

Weighted Bootstrap and Reweighting Methods

Resampling techniques using adaptive weights for improved efficiency in estimation and inference problems.

Spatial Econometrics and Regional Analysis

Statistical methods incorporating spatial dependence in economic data and regional development analysis.

Gaussian Process Regression and Kriging

Flexible nonparametric regression using Gaussian processes with uncertainty quantification and spatial interpolation.

Robust Principal Component Analysis

Methods for decomposing data into low-rank and sparse components robust to outliers and corruptions.

Sequential Decision Making Under Uncertainty

Statistical frameworks for optimal sequential decisions balancing exploration and exploitation in uncertain environments.

Causal Mediation Analysis and Pathway Analysis

Statistical methods for decomposing causal effects into direct and indirect pathways through mediators.

Functional Regression and Functional Principal Components

Regression and dimension reduction methods for functional predictors and responses in infinite-dimensional spaces.

Generative Adversarial Networks Theory

Statistical theory of adversarial training dynamics and convergence properties of generative models.

Doubly Robust Estimation and Augmented Approaches

Semiparametric methods that remain consistent if either outcome model or propensity score model is correctly specified.

Meta-Analysis and Evidence Synthesis

Statistical methods for combining results from multiple studies with heterogeneity assessment and publication bias.

Graphical Structure Learning from Data

Algorithms for discovering conditional independence structures and causal graphs from observational data.

Risk Stratification and Prediction Scoring

Development of statistical risk models for patient stratification and clinical prediction in healthcare.

Nonconvex Optimization in Statistics

Analysis of convergence and statistical properties of nonconvex optimization algorithms in learning problems.

Permutation Tests and Exact Inference

Distribution-free inference methods based on permutations applicable to complex and discrete data problems.

Survival Analysis with Interval Censoring

Statistical methods for event time analysis when exact event times are unknown but bounded by intervals.

Attention Mechanisms and Interpretability

Statistical analysis of attention-based neural networks with focus on model interpretability and explanation.

Nonlinear Dimension Reduction Techniques

Advanced methods for reducing dimensionality while preserving nonlinear structure in complex datasets.

Multiple Comparisons and Simultaneous Inference

Control of error rates in multiple hypothesis tests with methods for simultaneous confidence intervals.

Contaminated Data and Mixture Outliers

Statistical methods for analyzing datasets containing unknown proportions of outliers or contamination.

Latent Variable Models and Structural Equations

Comprehensive methods for modeling relationships among observed and latent variables in structural frameworks.

Survival Curves and Kaplan-Meier Estimation

Nonparametric estimation and inference for survival distributions under censoring mechanisms.

Convergence Rates and Lower Bounds Theory

Theoretical analysis of optimal convergence rates and fundamental statistical limits in estimation problems.

Cross-Validation and Model Assessment Techniques

Advanced methods for evaluating predictive performance and selecting among competing statistical models.

Causal Inference in Complex Networked Systems

Development of causal inference methods for interventions in interdependent network structures where traditional independence assumptions are violated.

Sequential Decision Making Under Model Misspecification

Theory and algorithms for adaptive sequential experiments when the underlying data-generating process may deviate from assumed models.

Moment-Based Inference and GMM Estimation

Generalized method of moments and moment-based approaches for parameter estimation in econometric models.

Adversarial Robustness and Perturbations

Statistical analysis of adversarial attacks and development of robust methods against adversarial perturbations.

Distributional Robustness and Worst-Case Optimization

Statistical methods for decision-making that maintain performance guarantees across ambiguous probability distributions within specified uncertainty sets.

Measurement Error and Misclassification Models

Methods for inference when covariates or outcomes are observed with measurement error or classification error.

Nonstationary Time Series and Structural Breaks

Estimation and inference procedures for time series with unknown change points and time-varying parameters in autoregressive and multivariate settings.

Latent Dirichlet Allocation Topic Modeling

Probabilistic methods for discovering latent topics in text corpora and high-dimensional document collections.

Statistical Methods for Single-Cell Genomics

Development of scalable statistical approaches for analyzing sparse, high-dimensional single-cell RNA-sequencing and protein expression data.

Causal Graphs and Backdoor Criterion

Graphical methods for identifying confounding paths and valid adjustment sets in causal inference.

Causal Mediation Analysis with Complex Mechanisms

Advanced methods for decomposing causal effects through multiple correlated mediators and interactions in observational and experimental studies.

Normalization and Batch Effect Correction

Statistical methods for removing technical variation and batch effects in high-throughput biological data.

Validation and Calibration of Predictive Models

Statistical techniques for assessing model reliability, identifying miscalibration, and developing valid uncertainty estimates in prediction tasks.

Information Geometry and Statistical Divergences

Applications of differential geometric concepts to understand statistical inference, model selection, and optimal transport between probability distributions.

Sequential Testing and Early Stopping

Methods for conducting hypothesis tests with data-dependent stopping times while controlling error rates.

Approximate Bayesian Inference for Intractable Likelihoods

Advanced computational methods including likelihood-free inference, ABC, and score-based approaches for models with analytically intractable likelihood functions.