Removing spurious features can hurt model accuracy and disproportionately affect different groups.
problem Interference from spurious features in robust model performance across different groups.
method Characterization and analysis of spurious feature removal in noiseless overparameterized linear regression.
result Removal of spurious features can decrease accuracy and disproportionately affect different groups, even in balanced datasets.
Unified framework for model explanation methods based on feature removal.
problem Unclear relationships and preferences among various model explanation methods.
method Characterizes removal-based explanations along three dimensions.
result Unified 26 existing methods, including widely used approaches.
Unified framework for explaining models by removing features.
problem Unclear relationships and preferences among existing model explanation methods.
method Removal-based explanations characterized by three dimensions.
result Unified framework unifies 26 existing methods.
Unified analysis of removal-based feature attributions robustness.
problem Robustness of removal-based feature attributions is not well understood.
method Theoretical analysis and upper bounds derivation for removal-based feature attributions under input and model perturbations.
result Upper bounds for the difference between intact and perturbed attributions derived under various perturbation settings.
Study removes bias from chest X-ray embeddings using orthogonalization.
problem Reduces bias in chest X-ray embeddings due to protected features.
method Orthogonalization technique to remove protected feature effects.
result Orthogonalization removes bias and makes predictions of protected attributes infeasible.
Unified theory explains how data augmentation improves deep learning models.
problem Understanding why data augmentation improves model generalization.
method Unified theoretical framework explaining two key effects: partial semantic feature removal and feature mixing.
result Data augmentation enhances generalization through partial semantic feature removal and feature mixing.
Perturbation-based explanation methods often measure the contribution of an input feature to an image classifier's outputs by heuristically removing it via e.g. blurring, adding noise, or graying out, which often produce unrealistic, out-of-samples. Instead, we propose to integrate a generative inpainter into three rep…
In this paper, we present a reverberation removal approach for speaker verification, utilizing dual-label deep neural networks (DNNs). The networks perform feature mapping between the spectral features of reverberant and clean speech. Long short term memory recurrent neural networks (LSTMs) are trained to map corrupted…
A new algorithm removes unexpected correlations in biased data for better clustering.
problem Clustering with selection bias in data.
method Decorrelation regularized K-Means (DCKM) algorithm.
result DCKM achieves significant performance gains on real-world datasets.
Online algorithm detects and removes spurious features for strong generalization.
problem Predictive models learn spurious correlations from incomplete data.
method Online continual learning algorithm that detects and removes spurious features.
result Algorithm converges to solutions with strong generalization.
New feature selection method DRPT reduces genomic datasets by removing irrelevant features and detecting correlations.
problem Feature selection in high-dimensional genomic datasets.
method DRPT method: 1. Remove irrelevant features, 2. Detect correlations in reduced matrix.
result DRPT outperforms state-of-the-art methods in feature selection over multiple genetic datasets.
Algorithm removes spurious concepts from neural network representations without harming task performance.
problem Spurious correlations hinder neural network out-of-distribution generalization.
method Iterative algorithm that identifies two orthogonal subspaces in neural network representation.
result Algorithm outperforms existing methods on computer vision and natural language processing benchmarks.
Optimized fuzzy entropy framework improves feature selection and classification performance.
problem Improving feature selection and classification in fuzzy entropy frameworks.
method Implemented and compared combinations of ideal vectors, maximal similarity classifiers, and fuzzy entropy functions.
result Optimized combination of ideal vector, similarity classifier, and fuzzy entropy function achieved the most stable performance for all three datasets.
Machine learning predicts S&P 500 additions and removals with high accuracy.
problem Forecasting S&P 500 membership changes to inform investment decisions.
method Used Random Forest model on quarterly data from 2013 onwards, incorporating various features.
result Achieved a test F1 score of 0.85, outperforming other models.
The paper tackles inconsistency in removal-based explanations and proposes methods to reduce it.
problem Inconsistency in removal-based explanations.
method Established the Impossible Trinity Theorem and proposed two novel algorithms to minimize interpretation error.
result The proposed methods achieve a substantial reduction in interpretation error, up to 31.8 times lower.
Unified framework for removing unwanted information from machine learning models.
problem Removing undesirable features or data points from machine learning models while preserving utility.
method Information-theoretic regularization approach for data point and feature unlearning.
result Unified mathematical framework with provable guarantees for both data point and feature unlearning.
UMFI improves feature importance methods by reducing runtime and enhancing performance.
problem Improving feature importance methods to better explain causal and associative relationships in data.
method Introducing UMFI, which uses dependence removal techniques from AI fairness literature.
result UMFI outperforms MCI, especially in complex data scenarios, and reduces runtime from exponential to super-linear.
Recent advances in Representation Learning and Adversarial Training seem to succeed in removing unwanted features from the learned representation. We show that demographic information of authors is encoded in -- and can be recovered from -- the intermediate representations learned by text-based neural classifiers. The …
Safe screening rules reduce computation time in logistic regression with ℓ0−ℓ2 regularization.
problem Efficiently solving logistic regression with many features and regularization.
method Screening rules based on Fenchel dual lower bounds of strong conic relaxations.
result A high percentage of features can be safely removed before solving, leading to substantial speed-up.
Debias concept-based explanations by removing confounding information.
problem Correlation between concepts and confounding features.
method Causal prior graph and two-stage regression technique.
result Success in removing biases and improving concept ranking.
Method matches noisy remote sensing images robustly.
problem Matching noisy remote sensing images.
method Combining attention mechanism with feature enhancement.
result More efficient and accurate matches achieved.
In this paper we introduce a new feature selection algorithm to remove the irrelevant or redundant features in the data sets. In this algorithm the importance of a feature is based on its fitting to the Catastrophe model. Akaike information crite- rion value is used for ranking the features in the data set. The propose…
This paper evaluates and improves metrics for identifying important features in machine learning models.
problem Evaluation metrics for explainable AI are limited by multicollinearity and model accuracy.
method Proposes Expected Accuracy Interval (EAI) to predict model accuracy with multicollinearity.
result EAI is a useful metric for identifying important features in models with multicollinearity.
Layer-wise Relevance Propagation (LRP) and saliency maps have been recently used to explain the predictions of Deep Learning models, specifically in the domain of text classification. Given different attribution-based explanations to highlight relevant words for a predicted class label, experiments based on word deleti…
Remove symmetries to improve model optimization and performance.
problem Symmetries in loss functions trap models in low-capacity states, hindering training and optimization.
method Proposes syre, a simple algorithm to remove symmetries in neural networks.
result Removing symmetries correlates well with improved optimization and performance.
Robust ASR model removes fast-changing features to resist attacks.
problem Vulnerability of ASR systems to adversarial attacks.
method Removing fast-changing features using slow feature analysis or low-pass filtering.
result Hybrid ASR models are more than four times more robust against targeted attacks.
Pruning method removes less important features in linear models.
problem Removing less important features in linear models trained by gradient flow.
method Iterative Magnitude Pruning (IMP) applied to linear models trained by gradient flow.
result IMP prunes features with smallest projection onto the data.
Bispectral OT improves dataset comparison by preserving intrinsic coherence.
problem Ignoring intrinsic coherence in dataset comparisons using pairwise geometric distances.
method Introduces Bispectral Optimal Transport, a symmetry-aware extension of discrete OT.
result Transport plans computed with Bispectral OT achieve greater class preservation accuracy.
Multiplicative noise, including dropout, is widely used to regularize deep neural networks (DNNs), and is shown to be effective in a wide range of architectures and tasks. From an information perspective, we consider injecting multiplicative noise into a DNN as training the network to solve the task with noisy informat…
Study explores DNNs' reliance on existing vs. new features in physiological signals.
problem Understanding how deep neural networks discover new features in physiological signals.
method Proposes a method to remove hand-engineered features and force DNNs to learn new representations.
result DNNs often rediscover known features, but can also learn new ones.
We focus in this paper on dataset reduction techniques for use in k-nearest neighbor classification. In such a context, feature and prototype selections have always been independently treated by the standard storage reduction algorithms. While this certifying is theoretically justified by the fact that each subproblem …
Dirichlet pruning compresses neural networks by removing unimportant units.
problem Compressing large neural network models without sacrificing performance.
method Assigns Dirichlet distribution over network layers' units and uses variational inference to estimate parameters.
result Achieves state-of-the-art compression performance on larger architectures like VGG and ResNet.
In text mining, information retrieval, and machine learning, text documents are commonly represented through variants of sparse Bag of Words (sBoW) vectors (e.g. TF-IDF). Although simple and intuitive, sBoW style representations suffer from their inherent over-sparsity and fail to capture word-level synonymy and polyse…
A new unsupervised method removes CT metal artifacts using beta-CycleGAN and attention.
problem Metal artifact reduction in computed tomography (CT) images.
method Unsupervised learning using a beta-CycleGAN architecture with attention mechanism.
result Improved metal artifact removal that preserves image details.
A new method guides neural networks to focus on specific features of input data.
problem Training neural networks to consider specific features of input data.
method Focus-and-Expand ( ax) method: Gradual manipulation of input features.
result Achieves state-of-the-art results in bias removal and image classification tasks.
An important research direction in machine learning has centered around developing meta-learning algorithms to tackle few-shot learning. An especially successful algorithm has been Model Agnostic Meta-Learning (MAML), a method that consists of two optimization loops, with the outer loop finding a meta-initialization, f…
Improved preterm prediction using synthetic EHG signals.
problem Prediction bias towards term labor in preterm EHG data.
method Quantifying synthetic samples' effect, optimizing feature weights, and combining activation functions.
result Substantial improvement in prediction precision.
Boilerplate removal refers to the problem of removing noisy content from a webpage such as ads and extracting relevant content that can be used by various services. This can be useful in several features in web browsers such as ad blocking, accessibility tools such as read out loud, translation, summarization etc. In o…
Study shows safely discarding features based on aggregate SHAP values is sound.
problem Discarding features based on aggregate SHAP values without proper justification.
method Investigated the soundness of discarding features based on aggregate SHAP values, proposing to aggregate SHAP values over the extended support.
result A small aggregate SHAP value implies safely discarding the corresponding feature.
Training features used to analyse physical processes are often highly correlated and determining which ones are most important for the classification is a non-trivial tasks. For the use case of a search for a top-quark pair produced in association with a Higgs boson decaying to bottom-quarks at the LHC, we compare feat…
Screening is an effective technique for speeding up the training process of a sparse learning model by removing the features that are guaranteed to be inactive the process. In this paper, we present a efficient screening technique for sparse support vector machine based on variational inequality. The technique is both …
Adversarial training purifies hidden weights to remove small perturbations.
problem Understanding and removing adversarial perturbations in deep learning models.
method Introducing Feature Purification, a principle that adversarial training aims to remove small dense mixtures in hidden weights.
result Adversarial training can make neural networks robust against small perturbations, even with simple algorithms.
Value selection reduces model size while maintaining accuracy.
problem Space efficiency in model size reduction.
method Two probabilistic methods based on information theory's metric: PVS and P + VS.
result Value selection achieves balance between accuracy and model size reduction.
Classic feature selection techniques remove those features that are either irrelevant or redundant, achieving a subset of relevant features that help to provide a better knowledge extraction. This allows the creation of compact models that are easier to interpret. Most of these techniques work over the whole dataset, b…
SIMPGEN improves SWOT SSH data interpretation by removing noise and preserving fine-scale features.
problem Noisy data and limited fine-scale observations in oceanic processes.
method Simulation-Informed Metric and Prior for Generative Ensemble Networks (SIMPGEN) combining real SWOT observations with simulated reference data.
result SIMPGEN effectively removes noise, preserving fine-scale features better than existing neural methods.
We propose a computationally efficient wrapper feature selection method - called Autoencoder and Model Based Elimination of features using Relevance and Redundancy scores (AMBER) - that uses a single ranker model along with autoencoders to perform greedy backward elimination of features. The ranker model is used to pri…
In matrix factorization, available graph side-information may not be well suited for the matrix completion problem, having edges that disagree with the latent-feature relations learnt from the incomplete data matrix. We show that removing these contested edges improves prediction accuracy and scalability. We…
We present a scheme for online, unsupervised state discovery and detection from streaming, multi-featured, asynchronous data in high-frequency financial markets. Online feature correlations are computed using an unbiased, lossless Fourier estimator. A high-speed maximum likelihood clustering algorithm is then used to f…