Corpus ID: 52891994

EDDI: Efficient Dynamic Discovery of High-Value Information with Partial VAE

@inproceedings{Ma2019EDDIED,
  title={EDDI: Efficient Dynamic Discovery of High-Value Information with Partial VAE},
  author={Chao Ma and Sebastian Tschiatschek and Konstantina Palla and Jos{\'e} Miguel Hern{\'a}ndez-Lobato and Sebastian Nowozin and Cui-cui Zhang},
  booktitle={ICML},
  year={2019}
}
Making decisions requires information relevant to the task at hand. [...] Key Method In EDDI we propose a novel partial variational autoencoder (Partial VAE), to efficiently handle missing data over varying subsets of known information. EDDI combines this Partial VAE with an acquisition function that maximizes expected information gain on a set of target variables. EDDI is efficient and demonstrates that dynamic discovery of high-value information is possible; we show cost reduction at the same decision quality…Expand
Icebreaker: Element-wise Active Information Acquisition with Bayesian Deep Latent Gaussian Model
TLDR
This paper proposes Icebreaker, a principled framework to approach the ice-start problem, a full Bayesian Deep Latent Gaussian Model (BELGAM) with a novel inference method that combines recent advances in amortized inference and stochastic gradient MCMC to enable fast and accurate posterior inference. Expand
Icebreaker: Element-wise Efficient Information Acquisition with a Bayesian Deep Latent Gaussian Model
TLDR
This paper proposes Icebreaker, a principled framework for elementwise training data acquisition, which introduces a full Bayesian Deep Latent Gaussian Model (BELGAM) with a novel inference method, which combines recent advances in amortized inference and stochastic gradient MCMC to enable fast and accurate posterior inference. Expand
FIT: a Fast and Accurate Framework for Solving Medical Inquiring and Diagnosing Tasks
TLDR
FIT improves over previous information-based approaches by using a multimodal variational autoencoder model and a two-step sampling strategy for disease prediction and is shown to be a competitive alternative in real-world settings. Expand
Reinforcement Learning with Efficient Active Feature Acquisition
TLDR
A model-based reinforcement learning framework that learns an active feature acquisition policy to solve the exploration-exploitation problem during its execution that outperforms conventional baselines and results in policies with greater cost efficiency. Expand
A Clustering based Selection Framework for Cost Aware and Test-time Feature Elicitation
TLDR
A Clustering based Cost Aware Test-time Feature Elicitation (CATE) algorithm, which can select the relevant feature set given the observed attributes of the test instance given that the cost of the features has already been paid in obtaining the training data. Expand
Conservative Policy Construction Using Variational Autoencoders for Logged Data with Missing Values
TLDR
This paper considers the problem of constructing personalized policies using logged data when there are missing values in the attributes of features in both training and test data, and introduces the conservative strategy where the policy is designed to safely handle the uncertainty due to missingness. Expand
Adaptive covariate acquisition for minimizing total cost of classification
TLDR
The proposed method achieves the target recall while minimizing the false discovery rate and the covariate acquisition costs better than previous methods, and is shown to achieve in most situations the lowest total costs when compared to various previous methods. Expand
ODIN: Optimal Discovery of High-value INformation Using Model-based Deep Reinforcement Learning
TLDR
A modelbased deep reinforcement learning framework for Optimal Discovery of high-value INformation (ODIN) in which the agent either chooses to ask for a new feature or to stop and predict is proposed, which ensures the globally optimal solution for most efficient feature acquisition while preserving data efficiency. Expand
Interpretable Outcome Prediction with Sparse Bayesian Neural Networks in Intensive Care
TLDR
This work proposes a novel interpretable Bayesian neural network architecture which offers both the flexibility of ANNs and interpretability in terms of feature selection, and employs a sparsity inducing prior distribution in a tied manner to learn which features are important for outcome prediction. Expand
Gradient Importance Learning for Incomplete Observations
  • Qitong Gao, Dong Wang, +6 authors M. Pajic
  • Computer Science
  • 2021
TLDR
The gradient importance learning (GIL) method is introduced to train multilayer perceptrons (MLPs) and long short-term memories (LSTMs) to directly perform inference from inputs containing missing values without imputation, and is employed to adjust the gradients used to train these models via back-propagation. Expand
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 41 REFERENCES
Active Feature-Value Acquisition
TLDR
This paper presents an acquisition policy, sampled expected utility (SEU), that employs particular estimations to enable effective ranking of potential acquisitions in settings where relatively little information is available about the underlying domain, and presents experimental results showing that SEU reduces the cost of producing a model of a desired accuracy and exhibits consistent performance across domains. Expand
Joint Active Feature Acquisition and Classification with Variable-Size Set Encoding
TLDR
This work forms this active feature acquisition as a jointly learning problem of training both the classifier (environment) and the RL agent that decides either to `stop and predict' or `collect a new feature' at test time, in a cost-sensitive manner. Expand
Active Feature Acquisition with Supervised Matrix Completion
TLDR
This paper tries to train an effective classification model with least acquisition cost by jointly performing active feature querying and supervised matrix completion and a bi-objective optimization method is presented for cost-aware active selection when features bear different acquisition costs. Expand
Handling Incomplete Heterogeneous Data using VAEs
TLDR
A general framework to design VAEs suitable for fitting incomplete heterogenous data, which includes likelihood models for real-valued, positive real valued, interval, categorical, ordinal and count data, and allows accurate estimation of missing data is proposed. Expand
Active feature-value acquisition for classifier induction
TLDR
This work presents an approach in which instances are selected for acquisition based on the current model's accuracy and its confidence in the prediction, which can induce accurate models using substantially fewer feature-value acquisitions as compared to alternative policies. Expand
On active learning for data acquisition
TLDR
This paper presents initial methods for data acquisition and evaluates these methods experimentally on web usage data and UCI datasets to indicate that active learning based methods forData acquisition can be a promising area for data mining research. Expand
Information-Based Objective Functions for Active Data Selection
  • D. Mackay
  • Mathematics, Computer Science
  • Neural Computation
  • 1992
TLDR
Within a Bayesian learning framework, objective functions are discussed that measure the expected informativeness of candidate measurements that depend on the assumption that the hypothesis space is correct. Expand
Bayesian Active Learning for Classification and Preference Learning
TLDR
This work proposes an approach that expresses information gain in terms of predictive entropies, and applies this method to the Gaussian Process Classier (GPC), and makes minimal approximations to the full information theoretic objective. Expand
Multitask learning and benchmarking with clinical time series data
TLDR
This work proposes four clinical prediction benchmarks using data derived from the publicly available Medical Information Mart for Intensive Care (MIMIC-III) database, covering a range of clinical problems including modeling risk of mortality, forecasting length of stay, detecting physiologic decline, and phenotype classification. Expand
Auto-Encoding Variational Bayes
TLDR
A stochastic variational inference and learning algorithm that scales to large datasets and, under some mild differentiability conditions, even works in the intractable case is introduced. Expand
...
1
2
3
4
5
...