Recurrent Neural Networks for Adaptive Feature Acquisition

Contardo, Gabriella; Denoyer, Ludovic; Artières, Thierry

doi:10.1007/978-3-319-46675-0_65

Cited by 17 publications

(16 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Based on the well trained random forest classifier, it prunes the RF by linear programming to balance between cost and accuracy [21]. RADIN: An adaptive learning method by using recurrent neural networks(RNN) with attention to select a fixed block of features and also fix the selection step [22]. DWSM:A method to formalize the task as an Markov decision process(MDP) and solve it with linearly approximated Qlearning [13].…”

Section: A Comparison Methodsmentioning

confidence: 99%

Costly Features Classification using Monte Carlo Tree Search

Chen

Huang

Ahn

et al. 2021

2021 International Joint Conference on Neural Networks (IJCNN)

View full text Add to dashboard Cite

In many real-world tasks, acquiring features requires a certain cost, which gives rise to the costly features classification problem. In this study, We formulate the problem in the reinforcement learning framework and sequentially select the subset of features to make a balance between the classification error and the feature cost. Specifically, advantage actor critic algorithm is firstly used to solve it. Furthermore, to improve the learned policy and make it explainable, we employ the Monte Carlo Tree Search to update the policy iteratively. During the procedure, we also consider its performance on imbalanced datasets. Our empirical evaluation shows that our method performs well in comparison with other traditional methods.

show abstract

Section: A Comparison Methodsmentioning

confidence: 99%

Costly Features Classification using Monte Carlo Tree Search

Chen

Huang

Ahn

et al. 2021

2021 International Joint Conference on Neural Networks (IJCNN)

View full text Add to dashboard Cite

show abstract

“…The model selection is made by training multiple models, selecting the best models on the validation set, and computing their performance on the test set. Note that the 'best models' in terms of both accuracy and FLOPs are the models located on the pareto front of the accuracy/cost validation curve as done for instance in [4]. These models are also evaluated using the matched, correct, wrong and false alarm (FA) metrics as proposed in [16] and computed over the one hour stream provided with the original dataset.…”

Section: Methodsmentioning

confidence: 99%

“…The highlighted architecture is the base model on which we have added shortcut connections. Conv1 and Conv2 have filter sizes of (20,8) and (10,4). Both have 64 channels and Conv1 has a stride of 3 in the frequency domain.…”

Section: Problem Definitionmentioning

confidence: 99%

Stochastic Adaptive Neural Architecture Search for Keyword Spotting

Véniat

Schwander

Denoyer

2019

ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

Self Cite

View full text Add to dashboard Cite

The problem of keyword spotting i.e. identifying keywords in a real-time audio stream is mainly solved by applying a neural network over successive sliding windows. Due to the difficulty of the task, baseline models are usually large, resulting in a high computational cost and energy consumption level. We propose a new method called SANAS (Stochastic Adaptive Neural Architecture Search) which is able to adapt the architecture of the neural network on-the-fly at inference time such that small architectures will be used when the stream is easy to process (silence, low noise, ...) and bigger networks will be used when the task becomes more difficult. We show that this adaptive model can be learned end-to-end by optimizing a trade-off between the prediction performance and the average computational cost per unit of time. Experiments on the Speech Commands dataset [16] show that this approach leads to a high recognition level while being much faster (and/or energy saving) than classical approaches where the network architecture is static.

show abstract

“…Figure 2: Evaluation of the proposed method on MNIST dataset. Accuracy vs. number of acquired features for OL, RADIN (Contardo et al, 2016), GreedyMiser , and a recent work based on reinforcement learning (RL-Based) (Janisch et al, 2017). , Cronus (Chen et al, 2012), andEarly Exit (Cambazoglu et al, 2010) approaches.…”

Section: Datasets and Experimentsmentioning

confidence: 99%

Opportunistic Learning: Budgeted Cost-Sensitive Learning from Data Streams

Kachuee,

Goldstein,

Karkkainen

et al. 2019

Preprint

View full text Add to dashboard Cite

In many real-world learning scenarios, features are only acquirable at a cost constrained under a budget. In this paper, we propose a novel approach for costsensitive feature acquisition at the prediction-time. The suggested method acquires features incrementally based on a context-aware feature-value function. We formulate the problem in the reinforcement learning paradigm, and introduce a reward function based on the utility of each feature. Specifically, MC dropout sampling is used to measure expected variations of the model uncertainty which is used as a feature-value function. Furthermore, we suggest sharing representations between the class predictor and value function estimator networks. The suggested approach is completely online and is readily applicable to stream learning setups. The solution is evaluated on three different datasets including the well-known MNIST dataset as a benchmark as well as two cost-sensitive datasets: Yahoo Learning to Rank and a dataset in the medical domain for diabetes classification. According to the results, the proposed method is able to efficiently acquire features and make accurate predictions.

show abstract

Recurrent Neural Networks for Adaptive Feature Acquisition

Cited by 17 publications

References 10 publications

Costly Features Classification using Monte Carlo Tree Search

Costly Features Classification using Monte Carlo Tree Search

Stochastic Adaptive Neural Architecture Search for Keyword Spotting

Opportunistic Learning: Budgeted Cost-Sensitive Learning from Data Streams

Contact Info

Product

Resources

About