Online Learning of Nonparametric Mixture Models via Sequential Variational Approximation

By Dahua Lin. 2013

71 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

1. Zheda Mai
2. Mengdi Xu
3. Ruiwen Li
4. Wenhao Ding
5. Jiacheng Zhu
6. Hyunwoo J. Kim
7. Emily B. Fox
8. Scott Sanner

Citations in the corpus

• Online Continual Learning in Image Classification: An Empirical Survey (2021, methodology): “Specifically, CN-DPM is comprised of a group of experts, where each expert is responsible for a subset of the 1[17] proposes two gradient-based methods, and we select the more efficient one with better performance, dubbed GSS-Greedy data and the group is expanded based on the Dirichlet Process Mixtu”
• Streaming Variational Inference for Bayesian Nonparametric Mixture Models (2014, methodology): “As such, our framework forms a theoretically justified and general-purpose scheme for BNP streaming inference, encompassing previous heuristic and model-specific approaches, and with a structure that enables insight into BNP inference via EP.”
• Online gradient-based mixtures for transfer modulation in meta-learning (2018, methodology): “A sequential approximation for nonparametric mixtures with a similar threshold was proposed in Lin (2013) and Tank et al. (2015), where variational Bayes was used instead of point estimation in a DPMM. Unlike traditional nonparametric algorithms, our model does not refine the cluster assignments of…”
• Task-Agnostic Online Reinforcement Learning with an Infinite Mixture of Gaussian Processes (2020, methodology): “Sequential variational inference [16] for DP mixtures has been recently integrated with DNNs for image classification [11].”
• Reconciling meta-learning and continual learning with online mixtures of tasks (2018, unknown): “A CRP prior over z j allows some probability to be assigned to a new mixture component while the task identities are inferred in a sequential manner, and has therefore been key to recent online and stochastic learning of the DPMM [31].”
• VariGrow: Variational Architecture Growing for Task-Agnostic Continual Learning based on Bayesian Novelty (2022, background): “However, the posterior distribution becomes intractable with large and complex datasets, and existing sequential variational approximations are not flexible enough to capture the complexity of these datasets (Blei & Jordan, 2006; Lin, 2013; Lee et al., 2020; Kessler et al., 2019).”
• A Scalable Hierarchical Gaussian Process Classifier (2019, background): “Recently, a stochastic variational approach for the general Dirichlet process mixture model has been proposed in [59].”