PyTorchKR Papers

최신 논문

총 848편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2203.12701

On Understanding the Influence of Controllable Factors with a Feature Attribution Algorithm: a Medical Case Study

Feature attribution XAI algorithms enable their users to gain insight into the underlying patterns of large datasets through their feature importance calculation. Existing feature attribution algorithms treat all features in a dataset homogeneously, which may lead to misinterpretation of consequences of changing feature values. In this work, we consider partitioning features into controllable and uncontrollable parts and propose the Controllable fActor Feature Attribution (CAFA) approach to compute the relative importance of controllable features. We carried out experiments applying CAFA to two existing datasets and our own COVID-19 non-pharmaceutical control measures dataset. Experimental results show that with CAFA, we are able to exclude influences from uncontrollable features in our explanation while keeping the full dataset for prediction.

arXiv:2203.08215공식 코드 · Papers with Code

Auto-Gait: Automatic Ataxia Risk Assessment with Computer Vision on Gait Task Videos

In this paper, we investigated whether we can 1) detect participants with ataxia-specific gait characteristics (risk-prediction), and 2) assess severity of ataxia from gait (severity-assessment) using computer vision. We created a dataset of 155 videos from 89 participants, 24 controls and 65 diagnosed with (or are pre-manifest) spinocerebellar ataxias (SCAs), performing the gait task of the Scale for the Assessment and Rating of Ataxia (SARA) from 11 medical sites located in 8 different states across the United States. We develop a computer vision pipeline to detect, track, and separate out the participants from their surroundings and construct several features from their body pose coordinates to capture gait characteristics like step width, step length, swing, stability, speed, etc. Our risk-prediction model achieves 83.06% accuracy and an 80.23% F1 score. Similarly, our severity-assessment model achieves a mean absolute error (MAE) score of 0.6225 and a Pearson's correlation coefficient score of 0.7268. Our models still performed competitively when evaluated on data from sites not used during training. Furthermore, through feature importance analysis, we found that our models associate wider steps, decreased walking speed, and increased instability with greater ataxia severity, which is consistent with previously established clinical knowledge. Our models create possibilities for remote ataxia assessment in non-clinical settings in the future, which could significantly improve accessibility of ataxia care. Furthermore, our underlying dataset was assembled from a geographically diverse cohort, highlighting its potential to further increase equity. The code used in this study is open to the public, and the anonymized body pose landmark dataset is also available upon request.

arXiv:2203.06648

The Yield Curve as a Recession Leading Indicator. An Application for Gradient Boosting and Random Forest

Most representative decision tree ensemble methods have been used to examine the variable importance of Treasury term spreads to predict US economic recessions with a balance of generating rules for US economic recession detection. A strategy is proposed for training the classifiers with Treasury term spreads data and the results are compared in order to select the best model for interpretability. We also discuss the use of SHapley Additive exPlanations (SHAP) framework to understand US recession forecasts by analyzing feature importance. Consistently with the existing literature we find the most relevant Treasury term spreads for predicting US economic recession and a methodology for detecting relevant rules for economic recession detection. In this case, the most relevant term spread found is 3 month to 6 month, which is proposed to be monitored by economic authorities. Finally, the methodology detected rules with high lift on predicting economic recession that can be used by these entities for this propose. This latter result stands in contrast to a growing body of literature demonstrating that machine learning methods are useful for interpretation comparing many alternative algorithms and we discuss the interpretation for our result and propose further research lines aligned with this work.

arXiv:2203.06487공식 코드 · Papers with Code

Evaluating Explainable AI on a Multi-Modal Medical Imaging Task: Can Existing Algorithms Fulfill Clinical Requirements?

Being able to explain the prediction to clinical end-users is a necessity to leverage the power of artificial intelligence (AI) models for clinical decision support. For medical images, a feature attribution map, or heatmap, is the most common form of explanation that highlights important features for AI models' prediction. However, it is unknown how well heatmaps perform on explaining decisions on multi-modal medical images, where each image modality or channel visualizes distinct clinical information of the same underlying biomedical phenomenon. Understanding such modality-dependent features is essential for clinical users' interpretation of AI decisions. To tackle this clinically important but technically ignored problem, we propose the modality-specific feature importance (MSFI) metric. It encodes clinical image and explanation interpretation patterns of modality prioritization and modality-specific feature localization. We conduct a clinical requirement-grounded, systematic evaluation using computational methods and a clinician user study. Results show that the examined 16 heatmap algorithms failed to fulfill clinical requirements to correctly indicate AI model decision process or decision quality. The evaluation and MSFI metric can guide the design and selection of XAI algorithms to meet clinical requirements on multi-modal explanation.

arXiv:2203.04483

Error-based Knockoffs Inference for Controlled Feature Selection

Recently, the scheme of model-X knockoffs was proposed as a promising solution to address controlled feature selection under high-dimensional finite-sample settings. However, the procedure of model-X knockoffs depends heavily on the coefficient-based feature importance and only concerns the control of false discovery rate (FDR). To further improve its adaptivity and flexibility, in this paper, we propose an error-based knockoff inference method by integrating the knockoff features, the error-based feature importance statistics, and the stepdown procedure together. The proposed inference procedure does not require specifying a regression model and can handle feature selection with theoretical guarantees on controlling false discovery proportion (FDP), FDR, or k-familywise error rate (k-FWER). Empirical evaluations demonstrate the competitive performance of our approach on both simulated and real data.

arXiv:2203.04289공식 코드 · Papers with Code

Self-supervised learning for analysis of temporal and morphological drug effects in cancer cell imaging data

In this work, we propose two novel methodologies to study temporal and morphological phenotypic effects caused by different experimental conditions using imaging data. As a proof of concept, we apply them to analyze drug effects in 2D cancer cell cultures. We train a convolutional autoencoder on 1M images dataset with random augmentations and multi-crops to use as feature extractor. We systematically compare it to the pretrained state-of-the-art models. We further use the feature extractor in two ways. First, we apply distance-based analysis and dynamic time warping to cluster temporal patterns of 31 drugs. We identify clusters allowing annotation of drugs as having cytotoxic, cytostatic, mixed or no effect. Second, we implement an adversarial/regularized learning setup to improve classification of 31 drugs and visualize image regions that contribute to the improvement. We increase top-3 classification accuracy by 8% on average and mine examples of morphological feature importance maps. We provide the feature extractor and the weights to foster transfer learning applications in biology. We also discuss utility of other pretrained models and applicability of our methods to other types of biomedical data.

arXiv:2203.02295

Evaluating Local Model-Agnostic Explanations of Learning to Rank Models with Decision Paths

Local explanations of learning-to-rank (LTR) models are thought to extract the most important features that contribute to the ranking predicted by the LTR model for a single data point. Evaluating the accuracy of such explanations is challenging since the ground truth feature importance scores are not available for most modern LTR models. In this work, we propose a systematic evaluation technique for explanations of LTR models. Instead of using black-box models, such as neural networks, we propose to focus on tree-based LTR models, from which we can extract the ground truth feature importance scores using decision paths. Once extracted, we can directly compare the ground truth feature importance scores to the feature importance scores generated with explanation techniques. We compare two recently proposed explanation techniques for LTR models when using decision trees and gradient boosting models on the MQ2008 dataset. We show that the explanation accuracy in these techniques can largely vary depending on the explained model and even which data point is explained.

arXiv:2203.02124

Abuse and Fraud Detection in Streaming Services Using Heuristic-Aware Machine Learning

This work presents a fraud and abuse detection framework for streaming services by modeling user streaming behavior. The goal is to discover anomalous and suspicious incidents and scale the investigation efforts by creating models that characterize the user behavior. We study the use of semi-supervised as well as supervised approaches for anomaly detection. In the semi-supervised approach, by leveraging only a set of authenticated anomaly-free data samples, we show the use of one-class classification algorithms as well as autoencoder deep neural networks for anomaly detection. In the supervised anomaly detection task, we present a so-called heuristic-aware data labeling strategy for creating labeled data samples. We carry out binary classification as well as multi-class multi-label classification tasks for not only detecting the anomalous samples but also identifying the underlying anomaly behavior(s) associated with each one. Finally, using a systematic feature importance study we provide insights into the underlying set of features that characterize different streaming fraud categories. To the best of our knowledge, this is the first paper to use machine learning methods for fraud and abuse detection in real-world scale streaming services.

arXiv:2203.01363

Faking feature importance: A cautionary tale on the use of differentially-private synthetic data

Synthetic datasets are often presented as a silver-bullet solution to the problem of privacy-preserving data publishing. However, for many applications, synthetic data has been shown to have limited utility when used to train predictive models. One promising potential application of these data is in the exploratory phase of the machine learning workflow, which involves understanding, engineering and selecting features. This phase often involves considerable time, and depends on the availability of data. There would be substantial value in synthetic data that permitted these steps to be carried out while, for example, data access was being negotiated, or with fewer information governance restrictions. This paper presents an empirical analysis of the agreement between the feature importance obtained from raw and from synthetic data, on a range of artificially generated and real-world datasets (where feature importance represents how useful each feature is when predicting a the outcome). We employ two differentially-private methods to produce synthetic data, and apply various utility measures to quantify the agreement in feature importance as this varies with the level of privacy. Our results indicate that synthetic data can sometimes preserve several representations of the ranking of feature importance in simple settings but their performance is not consistent and depends upon a number of factors. Particular caution should be exercised in more nuanced real-world settings, where synthetic data can lead to differences in ranked feature importance that could alter key modelling decisions. This work has important implications for developing synthetic versions of highly sensitive data sets in fields such as finance and healthcare.

arXiv:2203.01928공식 코드 · Papers with Code

Label-Free Explainability for Unsupervised Models

Unsupervised black-box models are challenging to interpret. Indeed, most existing explainability methods require labels to select which component(s) of the black-box's output to interpret. In the absence of labels, black-box outputs often are representation vectors whose components do not correspond to any meaningful quantity. Hence, choosing which component(s) to interpret in a label-free unsupervised/self-supervised setting is an important, yet unsolved problem. To bridge this gap in the literature, we introduce two crucial extensions of post-hoc explanation techniques: (1) label-free feature importance and (2) label-free example importance that respectively highlight influential features and training examples for a black-box to construct representations at inference time. We demonstrate that our extensions can be successfully implemented as simple wrappers around many existing feature and example importance methods. We illustrate the utility of our label-free explainability paradigm through a qualitative and quantitative comparison of representation spaces learned by various autoencoders trained on distinct unsupervised tasks.

arXiv:2202.12512

MUC-driven Feature Importance Measurement and Adversarial Analysis for Random Forest

The broad adoption of Machine Learning (ML) in security-critical fields demands the explainability of the approach. However, the research on understanding ML models, such as Random Forest (RF), is still in its infant stage. In this work, we leverage formal methods and logical reasoning to develop a novel model-specific method for explaining the prediction of RF. Our approach is centered around Minimal Unsatisfiable Cores (MUC) and provides a comprehensive solution for feature importance, covering local and global aspects, and adversarial sample analysis. Experimental results on several datasets illustrate the high quality of our feature importance measurement. We also demonstrate that our adversarial analysis outperforms the state-of-the-art method. Moreover, our method can produce a user-centered report, which helps provide recommendations in real-life applications.

arXiv:2202.11919

Threading the Needle of On and Off-Manifold Value Functions for Shapley Explanations

A popular explainable AI (XAI) approach to quantify feature importance of a given model is via Shapley values. These Shapley values arose in cooperative games, and hence a critical ingredient to compute these in an XAI context is a so-called value function, that computes the "value" of a subset of features, and which connects machine learning models to cooperative games. There are many possible choices for such value functions, which broadly fall into two categories: on-manifold and off-manifold value functions, which take an observational and an interventional viewpoint respectively. Both these classes however have their respective flaws, where on-manifold value functions violate key axiomatic properties and are computationally expensive, while off-manifold value functions pay less heed to the data manifold and evaluate the model on regions for which it wasn't trained. Thus, there is no consensus on which class of value functions to use. In this paper, we show that in addition to these existing issues, both classes of value functions are prone to adversarial manipulations on low density regions. We formalize the desiderata of value functions that respect both the model and the data manifold in a set of axioms and are robust to perturbation on off-manifold regions, and show that there exists a unique value function that satisfies these axioms, which we term the Joint Baseline value function, and the resulting Shapley value the Joint Baseline Shapley (JBshap), and validate the effectiveness of JBshap in experiments.

arXiv:2202.11204

Study of Feature Importance for Quantum Machine Learning Models

Predictor importance is a crucial part of data preprocessing pipelines in classical and quantum machine learning (QML). This work presents the first study of its kind in which feature importance for QML models has been explored and contrasted against their classical machine learning (CML) equivalents. We developed a hybrid quantum-classical architecture where QML models are trained and feature importance values are calculated from classical algorithms on a real-world dataset. This architecture has been implemented on ESPN Fantasy Football data using Qiskit statevector simulators and IBM quantum hardware such as the IBMQ Mumbai and IBMQ Montreal systems. Even though we are in the Noisy Intermediate-Scale Quantum (NISQ) era, the physical quantum computing results are promising. To facilitate current quantum scale, we created a data tiering, model aggregation, and novel validation methods. Notably, the feature importance magnitudes from the quantum models had a much higher variation when contrasted to classical models. We can show that equivalent QML and CML models are complementary through diversity measurements. The diversity between QML and CML demonstrates that both approaches can contribute to a solution in different ways. Within this paper we focus on Quantum Support Vector Classifiers (QSVC), Variational Quantum Circuit (VQC), and their classical counterparts. The ESPN and IBM fantasy football Trade Assistant combines advanced statistical analysis with the natural language processing of Watson Discovery to serve up personalized trade recommendations that are fair. Here, player valuation data of each player has been considered and this work can be extended to calculate the feature importance of other QML models such as Quantum Boltzmann machines.

arXiv:2202.10453공식 코드 · Papers with Code

Predicting emotion from music videos: exploring the relative contribution of visual and auditory information to affective responses

Although media content is increasingly produced, distributed, and consumed in multiple combinations of modalities, how individual modalities contribute to the perceived emotion of a media item remains poorly understood. In this paper we present MusicVideos (MuVi), a novel dataset for affective multimedia content analysis to study how the auditory and visual modalities contribute to the perceived emotion of media. The data were collected by presenting music videos to participants in three conditions: music, visual, and audiovisual. Participants annotated the music videos for valence and arousal over time, as well as the overall emotion conveyed. We present detailed descriptive statistics for key measures in the dataset and the results of feature importance analyses for each condition. Finally, we propose a novel transfer learning architecture to train Predictive models Augmented with Isolated modality Ratings (PAIR) and demonstrate the potential of isolated modality ratings for enhancing multimodal emotion recognition. Our results suggest that perceptions of arousal are influenced primarily by auditory information, while perceptions of valence are more subjective and can be influenced by both visual and auditory information. The dataset is made publicly available.

arXiv:2202.02389

The impact of feature importance methods on the interpretation of defect classifiers

Classifier specific (CS) and classifier agnostic (CA) feature importance methods are widely used (often interchangeably) by prior studies to derive feature importance ranks from a defect classifier. However, different feature importance methods are likely to compute different feature importance ranks even for the same dataset and classifier. Hence such interchangeable use of feature importance methods can lead to conclusion instabilities unless there is a strong agreement among different methods. Therefore, in this paper, we evaluate the agreement between the feature importance ranks associated with the studied classifiers through a case study of 18 software projects and six commonly used classifiers. We find that: 1) The computed feature importance ranks by CA and CS methods do not always strongly agree with each other. 2) The computed feature importance ranks by the studied CA methods exhibit a strong agreement including the features reported at top-1 and top-3 ranks for a given dataset and classifier, while even the commonly used CS methods yield vastly different feature importance ranks. Such findings raise concerns about the stability of conclusions across replicated studies. We further observe that the commonly used defect datasets are rife with feature interactions and these feature interactions impact the computed feature importance ranks of the CS methods (not the CA methods). We demonstrate that removing these feature interactions, even with simple methods like CFS improves agreement between the computed feature importance ranks of CA and CS methods. In light of our findings, we provide guidelines for stakeholders and practitioners when performing model interpretation and directions for future research, e.g., future research is needed to investigate the impact of advanced feature interaction removal methods on computed feature importance ranks of different CS methods.

arXiv:2202.02298

Towards a consistent interpretation of AIOps models

Artificial Intelligence for IT Operations (AIOps) has been adopted in organizations in various tasks, including interpreting models to identify indicators of service failures. To avoid misleading practitioners, AIOps model interpretations should be consistent (i.e., different AIOps models on the same task agree with one another on feature importance). However, many AIOps studies violate established practices in the machine learning community when deriving interpretations, such as interpreting models with suboptimal performance, though the impact of such violations on the interpretation consistency has not been studied. In this paper, we investigate the consistency of AIOps model interpretation along three dimensions: internal consistency, external consistency, and time consistency. We conduct a case study on two AIOps tasks: predicting Google cluster job failures, and Backblaze hard drive failures. We find that the randomness from learners, hyperparameter tuning, and data sampling should be controlled to generate consistent interpretations. AIOps models with AUCs greater than 0.75 yield more consistent interpretation compared to low-performing models. Finally, AIOps models that are constructed with the Sliding Window or Full History approaches have the most consistent interpretation with the trends presented in the entire datasets. Our study provides valuable guidelines for practitioners to derive consistent AIOps model interpretation.

arXiv:2201.13378

Novel Machine Learning Approaches for Improving the Reproducibility and Reliability of Functional and Effective Connectivity from Functional MRI

Objective: New measures of human brain connectivity are needed to address gaps in the existing measures and facilitate the study of brain function, cognitive capacity, and identify early markers of human disease. Traditional approaches to measure functional connectivity between pairs of brain regions in functional MRI, such as correlation and partial correlation, fail to capture nonlinear aspects in the regional associations. We propose a new machine learning based measure of functional connectivity which efficiently captures linear and nonlinear aspects. Approach: We propose two new EC measures. The first, a machine learning based measure of effective connectivity, measures nonlinear aspects across the entire brain. The second, Structurally Projected Granger Causality adapts Granger Causal connectivity to efficiently characterize and regularize the whole brain EC connectome to respect underlying biological structural connectivity. The proposed measures are compared to traditional measures in terms of reproducibility and the ability to predict individual traits in order to demonstrate these measures internal validity. We use four repeat scans of the same individuals from the Human Connectome Project and measure the ability of the measures to predict individual subject physiologic and cognitive traits. Main results: The proposed new FC measure of ML.FC attains high reproducibility with an R squared of 0.44, while the proposed EC measure of SP.GC attains the highest predictive power with an R squared of 0.66. Significance: The proposed methods are highly suitable for achieving high reproducibility and predictiveness.

arXiv:2201.12380공식 코드 · Papers with Code

GStarX: Explaining Graph Neural Networks with Structure-Aware Cooperative Games

Explaining machine learning models is an important and increasingly popular area of research interest. The Shapley value from game theory has been proposed as a prime approach to compute feature importance towards model predictions on images, text, tabular data, and recently graph neural networks (GNNs) on graphs. In this work, we revisit the appropriateness of the Shapley value for GNN explanation, where the task is to identify the most important subgraph and constituent nodes for GNN predictions. We claim that the Shapley value is a non-ideal choice for graph data because it is by definition not structure-aware. We propose a Graph Structure-aware eXplanation (GStarX) method to leverage the critical graph structure information to improve the explanation. Specifically, we define a scoring function based on a new structure-aware value from the cooperative game theory proposed by Hamiache and Navarro (HN). When used to score node importance, the HN value utilizes graph structures to attribute cooperation surplus between neighbor nodes, resembling message passing in GNNs, so that node importance scores reflect not only the node feature importance, but also the node structural roles. We demonstrate that GStarX produces qualitatively more intuitive explanations, and quantitatively improves explanation fidelity over strong baselines on chemical graph property prediction and text graph sentiment classification.

arXiv:2201.12114공식 코드 · Papers with Code

Rethinking Attention-Model Explainability through Faithfulness Violation Test

Attention mechanisms are dominating the explainability of deep models. They produce probability distributions over the input, which are widely deemed as feature-importance indicators. However, in this paper, we find one critical limitation in attention explanations: weakness in identifying the polarity of feature impact. This would be somehow misleading -- features with higher attention weights may not faithfully contribute to model predictions; instead, they can impose suppression effects. With this finding, we reflect on the explainability of current attention-based techniques, such as Attentio$\odot$Gradient and LRP-based attention explanations. We first propose an actionable diagnostic methodology (henceforth faithfulness violation test) to measure the consistency between explanation weights and the impact polarity. Through the extensive experiments, we then show that most tested explanation methods are unexpectedly hindered by the faithfulness violation issue, especially the raw attention. Empirical analyses on the factors affecting violation issues further provide useful observations for adopting explanation methods in attention models.

arXiv:2201.11194

Attention cannot be an Explanation

Attention based explanations (viz. saliency maps), by providing interpretability to black box models such as deep neural networks, are assumed to improve human trust and reliance in the underlying models. Recently, it has been shown that attention weights are frequently uncorrelated with gradient-based measures of feature importance. Motivated by this, we ask a follow-up question: "Assuming that we only consider the tasks where attention weights correlate well with feature importance, how effective are these attention based explanations in increasing human trust and reliance in the underlying models?". In other words, can we use attention as an explanation? We perform extensive human study experiments that aim to qualitatively and quantitatively assess the degree to which attention based explanations are suitable in increasing human trust and reliance. Our experiment results show that attention cannot be used as an explanation.