CDLEEDS detects local changes in evolving data streams for accurate feature attributions.
problem Local feature attributions become obsolete in evolving data streams.
method CDLEEDS, a flexible framework for detecting local change and concept drift.
result CDLEEDS reliably detects both local and global concept drift.
Proposes methods for local clustering in attributed graphs.
problem Finding a single cluster concentrated on a specific region in a graph.
method Introduces Graph Unimodality (GU) and Attribute Unimodality (AU) measures, and LOCLU algorithm to optimize Compactness score.
result Local cluster detected by LOCLU concentrates on the region of interest and exhibits unimodal data distribution.
New method R-LOCO improves local feature importance analysis.
problem Local attribution methods fail to accurately identify important features.
method R-LOCO segments input space into regions and applies global methods within.
result R-LOCO delivers more accurate local attributions.
Local explanation methods, also known as attribution methods, attribute a deep network's prediction to its input (cf. Baehrens et al. (2010)). We respond to the claim from Adebayo et al. (2018) that local explanation methods lack sensitivity, i.e., DNNs with randomly-initialized weights produce explanations that are bo…
New findings show local attributions can't be both robust and provide recourse.
problem Ensuring machine learning systems are accountable and provide actionable recourse options.
method Formal definition of recourse sensitivity and counterexamples for popular attribution methods.
result It is impossible for any single attribution method to be both robust and provide recourse.
Improved local feature attributions using neighbourhood reference distributions.
problem Misleading results from global population in local model behaviour.
method Formulation of neighbourhood reference distributions and self-normalised importance sampling.
result Neighbourhood Shapley values provide meaningful sparse feature attributions.
Local decision boundary approximation improves model explanations for complex models.
problem Challenges in explaining complex, opaque machine learning models.
method Train a variational autoencoder to learn a latent space and map it to meaningful attributes. Use these attributes to approximate the local decision boundary and explain model predictions.
result Can recover latent attributes that determine class decisions in a new benchmark data set.
TaylorPODA uses Taylor expansions to improve feature attributions for opaque models.
problem Lack of systematic framework for quantifying feature contributions in opaque models.
method Taylor expansion framework with postulates (precision, federation, zero-discrepancy, adaptation).
result TaylorPODA achieves competitive results and provides principled explanations.
Adding attributes for nodes to network embedding helps to improve the ability of the learned joint representation to depict features from topology and attributes simultaneously. Recent research on the joint embedding has exhibited a promising performance on a variety of tasks by jointly embedding the two spaces. Howeve…
Unified analysis of multi-attribute graph learning with non-convex penalties.
problem Graph inference from multi-attribute data.
method Penalized log-likelihood objective function with ADMM and local linear approximation.
result Local consistency in support recovery and precision matrix estimation for non-convex penalties.
Extends local attributions to Bayesian Neural Networks for improved explanations.
problem Lack of explanations for Bayesian Neural Networks' predictions.
method Extend local attributions to a probabilistic explanation distribution of BNNs.
result Enriches standard explanations with uncertainty information and visualizes explanation stability.
Current approaches for explaining machine learning models fall into two distinct classes: antecedent event influence and value attribution. The former leverages training instances to describe how much influence a training point exerts on a test point, while the latter attempts to attribute value to the features most pe…
Wavelet Attribution Method (WAM) improves feature attribution for deep models.
problem Inability of pixel-based heatmaps to capture data structure and variability in feature attribution.
method Wavelet domain for feature attribution, leveraging spatial and scale-localized properties of wavelet coefficients.
result WAM provides quantitatively superior explanations across audio, image, and volume modalities.
Proposes VCLANC for attributed network clustering using node and attribute embeddings.
problem Lack of mutual affinity exploitation between nodes and attributes in graph convolution.
method Dual variational auto-encoders for node and attribute embeddings, Gaussian mixture model priors, mutual distance and clustering assignment hardening losses.
result Demonstrates effectiveness on real-world attributed network datasets.
MACQ method explains deep learning models by analyzing feature contributions across prediction levels.
problem Explaining deep learning model predictions.
method Global gradient-based, model-agnostic approach focusing on marginal attribution.
result MACQ separates feature contributions from interaction effects and visualizes 3-way relationships.
A new method assigns anomaly scores to features for better interpretation.
problem Interpreting anomaly scores from feature attributions.
method Proposes a characteristic function to attribute anomaly scores using Shapley value.
result Demonstrates the potential utility of the proposed attribution methods.
We present network embedding algorithms that capture information about a node from the local distribution over node attributes around it, as observed over random walks following an approach similar to Skip-gram. Observations from neighborhoods of different sizes are either pooled (AE) or encoded distinctly in a multi-s…
Improved 3D ECG feature attributions for clinical interpretation.
problem Lack of interpretability in deep learning models for 12-lead ECG analysis.
method Cross-modal mapping of feature attributions from 12-lead ECG models onto CineECG 3D space.
result Mapped feature attributions yield higher Dice scores than standard 12-lead attributions.
Proposes a method for explaining black-box models with nested feature attributions.
problem Making black-box models transparent and trustworthy.
method Model-agnostic local explanation method exploiting nested feature structure and consistency property.
result Accurate and consistent HiFAs and LoFAs estimated using fewer model queries.
Optimal LDP mechanisms reduce data unfairness in classification.
problem Reducing data unfairness in classification models.
method Developed a closed-form optimal mechanism for binary attributes and a tractable framework for multi-valued attributes.
result Optimal LDP mechanisms improve fairness in classification while maintaining accuracy close to non-private models.
PatternLocal improves XAI for non-linear models by suppressing suppressor variables.
problem Suppressor variables cause false-positive feature attributions in non-linear models.
method PatternLocal uses locally linear surrogate models and transforms weights into a generative representation.
result PatternLocal reduces false-positive attributions and provides more reliable explanations.
Unified framework for linear attribution methods in deep learning.
problem Separate theoretical foundations of XAI attribution methods.
method GRALIS (Gradient-Riesz Averaged Locally-Integrated Shapley) framework.
result Unified representation theory for linear attribution methods.
PredDiff measures prediction changes while marginalizing features, offering new insights into interaction effects.
problem Understanding interaction effects in black-box models.
method Model-agnostic, local attribution method based on probability theory.
result Introduced a new measure for interaction effects between arbitrary feature subsets.
Interpreting predictions from tree ensemble methods such as gradient boosting machines and random forests is important, yet feature attribution for trees is often heuristic and not individualized for each prediction. Here we show that popular feature attribution methods are inconsistent, meaning they can lower a featur…
Graph Convolutional Networks (GCNs) have proved to be a most powerful architecture in aggregating local neighborhood information for individual graph nodes. Low-rank proximities and node features are successfully leveraged in existing GCNs, however, attributes that graph links may carry are commonly ignored, as almost …
Proposes a new graph kernel framework using regularized Wasserstein distances.
problem Learning optimal transport distances for graph kernels.
method Introduces Regularized Wasserstein (RW) discrepancy with two regularization terms.
result Empirically validated method outperforms state-of-the-art methods.
Attributed network embedding has attracted plenty of interest in recent years. It aims to learn task-independent, low-dimensional, and continuous vectors for nodes preserving both topology and attribute information. Most of the existing methods, such as random-walk based methods and GCNs, mainly focus on the local info…
Existing techniques to compress point cloud attributes leverage either geometric or video-based compression tools. We explore a radically different approach inspired by recent advances in point cloud representation learning. Point clouds can be interpreted as 2D manifolds in 3D space. Specifically, we fold a 2D grid on…
SX-GeoTree improves spatially coherent explanations in geospatial regression trees.
problem Capturing spatial dependence and producing robust explanations in tabular prediction models.
method Integrates three objectives: impurity reduction, spatial residual control, and explanation robustness via modularity maximization on a consensus similarity network.
result Improves residual spatial evenness and doubles attribution consensus (modularity: Fujian 0.19 vs 0.09; Seattle 0.10 vs 0.05).
Tree Mover's Distance measures graph attributes and improves GNN performance.
problem Measuring generalization and robustness in graph neural networks.
method Introducing Tree Mover's Distance (TMD) for attributed graphs.
result TMD correlates with GNN performance under distribution shifts.
The widespread use of automated decision processes in many areas of our society raises serious ethical issues concerning the fairness of the process and the possible resulting discriminations. In this work, we propose a novel approach called GANsan whose objective is to prevent the possibility of any discrimination i.e…
Unified analysis for graph learning from multi-attribute Gaussian time series.
problem Estimating conditional independence graph from multi-attribute Gaussian time series data.
method Unified theoretical analysis using a penalized log-likelihood objective function in the frequency domain.
result Established sufficient conditions for consistency and graph recovery in high-dimensional settings.
This paper introduces a new method to cluster qualitative attribute data using tree structures.
problem Clustering qualitative attribute data, especially when values are not in Euclidean space.
method Developed a joint learning mechanism to iteratively learn trees representing qualitative values' order relationships.
result The joint learning mechanism successfully clusters qualitative attribute data, yielding accurate results.
A barrier to the wider adoption of neural networks is their lack of interpretability. While local explanation methods exist for one prediction, most global attributions still reduce neural network decisions to a single set of features. In response, we present an approach for generating global attributions called GAM, w…
Paper defends sensitive attributes in GNNs from inference attacks.
problem Protecting sensitive attributes in GNNs from inference attacks.
method Proposes adversarial training with TV and Wasserstein distance to locally filter sensitive attributes.
result Framework creates strong defense against inference attacks with minimal performance loss.
The paper introduces COAR to estimate component attributions and enable model editing.
problem Understanding how ML models transform inputs into predictions.
method Component modeling and COAR algorithm for estimating component attributions.
result COAR enables model editing across various tasks.
In healthcare, making the best possible predictions with complex models (e.g., neural networks, ensembles/stacks of different models) can impact patient welfare. In order to make these complex models explainable, we present DeepSHAP for mixed model types, a framework for layer wise propagation of Shapley values that bu…
The rise of algorithmic decision making led to active researches on how to define and guarantee fairness, mostly focusing on one-shot decision making. In several important applications such as hiring, however, decisions are made in multiple stage with additional information at each stage. In such cases, fairness issues…
Complex models are commonly used in predictive modeling. In this paper we present R packages that can be used to explain predictions from complex black box models and attribute parts of these predictions to input features. We introduce two new approaches and corresponding packages for such attribution, namely live and …
We show new connections between adversarial learning and explainability for deep neural networks (DNNs). One form of explanation of the output of a neural network model in terms of its input features, is a vector of feature-attributions. Two desirable characteristics of an attribution-based explanation are: (1) $\texti…
This work addresses local fairness in machine learning models.
problem Ensuring fairness within subregions of feature space, not just global averages.
method Introduces ROAD, a Distributionally Robust Optimization (DRO) approach with adversarial learning.
result Achieves Pareto dominance in local fairness and accuracy across datasets.
NetRCA algorithm locates network faults by analyzing derived features and leveraging unlabeled data.
problem Locating the true root cause of network faults is challenging due to complex architectures and limited labeled data.
method NetRCA algorithm extracts derived features, generates new training data, and combines multiple models to enhance performance.
result NetRCA outperforms existing methods in fault cause localization on real-world data.
For some or all of the data instances a number of independent-world clustering issues suffer from incomplete data characterization due to losing or absent attributes. Typical clustering approaches cannot be applied directly to such data unless pre-processing by techniques like imputation or marginalization. We have ove…
We introduce a comprehensive and statistical framework in a model free setting for a complete treatment of localized data corruptions due to severe noise sources, e.g., an occluder in the case of a visual recording. Within this framework, we propose i) a novel algorithm to efficiently separate, i.e., detect and localiz…
TSInsight improves interpretability of deep time-series models.
problem Lack of interpretability methods for time-series data.
method Attach auto-encoder to classifier with sparsity-inducing norm, fine-tune based on gradients and reconstruction penalty.
result TSInsight effectively boosts interpretability of deep time-series models.
New method discovers concepts in hidden feature layers using sparse subspace clustering.
problem Local attribution methods fail to identify coherent model behavior across samples.
method Sparse Subspace Clustering (SSCC) for concept discovery.
result Empirically validated method for various image classification tasks.
Note that a newer expanded version of this paper is now available at: arXiv:1802.03888 It is critical in many applications to understand what features are important for a model, and why individual predictions were made. For tree ensemble methods these questions are usually answered by attributing importance values to i…
RKHS-SHAP uses Shapley values for kernel methods to provide feature attributions.
problem Feature attribution for kernel methods is often heuristic and not individualised.
method RKHS-SHAP uses Shapley values from coalition game theory to compute feature attributions for kernel machines efficiently.
result RKHS-SHAP can compute both Interventional and Observational Shapley values.