Dirichlet Prior

The Dirichlet prior is a probability distribution used in Bayesian statistics to model uncertainty, particularly in scenarios involving categorical or count data. Current research focuses on its application in diverse fields, including clustering, causal inference, and uncertainty quantification in machine learning models, often employing it within Bayesian frameworks or as a component of novel algorithms like those incorporating auxiliary uncertainty estimators or large language models. This versatile tool enhances model robustness and allows for more nuanced interpretations of uncertainty, leading to improved decision-making in various applications, from medical diagnosis to data stream analysis.

Papers

August 2, 2024

Point Prediction for Streaming Data
Aleena Chanda, N. V. Vinodchandran, Bertrand Clarke
Streaming Data Point Prediction Dirichlet Prior Prediction Problem

February 3, 2024

A Bayesian cluster validity index
Nathakhun Wiroonsri, Onthada Preedasawakul
Hard Clustering Clustering Validation Dirichlet Prior

November 13, 2023

Applying Large Language Models for Causal Structure Learning in Non Small Cell Lung Cancer
Narmada Naik, Ayush Khandelwal, Mohit Joshi, Madhusudan Atre, Hollis Wright, Kavya Kannan, Scott Hill, Giridhar Mamidipudi, Ganapati Srinivasa, Carlo Bifulco, Brian Piening, Kevin Matlock
Causal Discovery Causal Graph Causal Relationship Medical AI Lung Cancer Causal Structure Learning Dirichlet Prior

August 17, 2023

Discretization-Induced Dirichlet Posterior for Robust Uncertainty Quantification on Regression
Xuanlong Yu, Gianni Franchi, Jindong Gu, Emanuel Aldea
Native Robustness Uncertainty Quantification Novel Regression Auxiliary Objective Robust Uncertainty Dirichlet Prior

March 8, 2022

Estimating the Uncertainty in Emotion Class Labels with Utterance-Specific Dirichlet Priors
Wen Wu, Chao Zhang, Xixin Wu, Philip C. Woodland
Emotion Recognition High Uncertainty Anticipation Discrete Emotion Emotion Class Dirichlet Prior

November 18, 2021

From Optimality to Robustness: Dirichlet Sampling Strategies in Stochastic Bandits
Dorian Baudry, Patrick Saux, Odalric-Ambrym Maillard
Native Robustness Multi Armed Bandit Near Optimality Regret Guarantee Stochastic Bandit Logarithmic Regret Logarithmic Bayes Regret Dirichlet Prior

Dirichlet Prior

Papers

Point Prediction for Streaming Data

A Bayesian cluster validity index

Applying Large Language Models for Causal Structure Learning in Non Small Cell Lung Cancer

Discretization-Induced Dirichlet Posterior for Robust Uncertainty Quantification on Regression

Estimating the Uncertainty in Emotion Class Labels with Utterance-Specific Dirichlet Priors

From Optimality to Robustness: Dirichlet Sampling Strategies in Stochastic Bandits