Substantial progress in spoofing and deepfake detection has been made in recent years. Nonetheless, the community has yet to make notable inroads in providing an explanation for how a classifier produces its output. The dominance of black box spoofing detection solutions is at further odds with the drive toward trustworthy, explainable artificial intelligence. This paper describes our use of SHapley Additive exPlanations (SHAP) to gain new insights in spoofing detection. We demonstrate use of the tool in revealing unexpected classifier behaviour, the artefacts that contribute most to classifier outputs and differences in the behaviour of competing spoofing detection models. The tool is both efficient and flexible, being readily applicable to a host of different architecture models in addition to related, different applications. All results reported in the paper are reproducible using open-source software.
While Explainable Artificial Intelligence (XAI) is increasingly expanding more areas of application, little has been applied to make deep Reinforcement Learning (RL) more comprehensible. As RL becomes ubiquitous and used in critical and general public applications, it is essential to develop methods that make it better understood and more interpretable. This study proposes a novel approach to explain cooperative strategies in multiagent RL using Shapley values, a game theory concept used in XAI that successfully explains the rationale behind decisions taken by Machine Learning algorithms. Through testing common assumptions of this technique in two cooperation-centered socially challenging multi-agent environments environments, this article argues that Shapley values are a pertinent way to evaluate the contribution of players in a cooperative multi-agent RL context. To palliate the high overhead of this method, Shapley values are approximated using Monte Carlo sampling. Experimental results on Multiagent Particle and Sequential Social Dilemmas show that Shapley values succeed at estimating the contribution of each agent. These results could have implications that go beyond games in economics, (e.g., for non-discriminatory decision making, ethical and responsible AI-derived decisions or policy making under fairness constraints). They also expose how Shapley values only give general explanations about a model and cannot explain a single run, episode nor justify precise actions taken by agents. Future work should focus on addressing these critical aspects.
Post-hoc explanation methods, e.g., Grad-CAM, enable humans to inspect the spatial regions responsible for a particular network decision. However, it is shown that such explanations are not always consistent with human priors, such as consistency across image transformations. Given an interpretation algorithm, e.g., Grad-CAM, we introduce a novel training method to train the model to produce more consistent explanations. Since obtaining the ground truth for a desired model interpretation is not a well-defined task, we adopt ideas from contrastive self-supervised learning, and apply them to the interpretations of the model rather than its embeddings. We show that our method, Contrastive Grad-CAM Consistency (CGC), results in Grad-CAM interpretation heatmaps that are more consistent with human annotations while still achieving comparable classification accuracy. Moreover, our method acts as a regularizer and improves the accuracy on limited-data, fine-grained classification settings. In addition, because our method does not rely on annotations, it allows for the incorporation of unlabeled data into training, which enables better generalization of the model. Our code is available here: https://github.com/UCDvision/CGC
AI explainability improves the transparency of models, making them more trustworthy. Such goals are motivated by the emergence of deep learning models, which are obscure by nature; even in the domain of images, where deep learning has succeeded the most, explainability is still poorly assessed. In the field of image recognition many feature attribution methods have been proposed with the purpose of explaining a model's behavior using visual cues. However, no metrics have been established so far to assess and select these methods objectively. In this paper we propose a consistent evaluation score for feature attribution methods -- the Focus -- designed to quantify their coherency to the task. While most previous work adds out-of-distribution noise to samples, we introduce a methodology to add noise from within the distribution. This is done through mosaics of instances from different classes, and the explanations these generate. On those, we compute a visual pseudo-precision metric, Focus. First, we show the robustness of the approach through a set of randomization experiments. Then we use Focus to compare six popular explainability techniques across several CNN architectures and classification datasets. Our results find some methods to be consistently reliable (LRP, GradCAM), while others produce class-agnostic explanations (SmoothGrad, IG). Finally we introduce another application of Focus, using it for the identification and characterization of biases found in models. This empowers bias-management tools, in another small step towards trustworthy AI.
Counterfactual explanation methods interpret the outputs of a machine learning model in the form of "what-if scenarios" without compromising the fidelity-interpretability trade-off. They explain how to obtain a desired prediction from the model by recommending small changes to the input features, aka recourse. We believe an actionable recourse should be created based on sound counterfactual explanations originating from the distribution of the ground-truth data and linked to the domain knowledge. Moreover, it needs to preserve the coherency between changed/unchanged features while satisfying user/domain-specified constraints. This paper introduces CARE, a modular explanation framework that addresses the model- and user-level desiderata in a consecutive and structured manner. We tackle the existing requirements by proposing novel and efficient solutions that are formulated in a multi-objective optimization framework. The designed framework enables including arbitrary requirements and generating counterfactual explanations and actionable recourse by choice. As a model-agnostic approach, CARE generates multiple, diverse explanations for any black-box model in tabular classification and regression settings. Several experiments on standard data sets and black-box models demonstrate the effectiveness of our modular framework and its superior performance compared to the baselines.
The large and still increasing popularity of deep learning clashes with a major limit of neural network architectures, that consists in their lack of capability in providing human-understandable motivations of their decisions. In situations in which the machine is expected to support the decision of human experts, providing a comprehensible explanation is a feature of crucial importance. The language used to communicate the explanations must be formal enough to be implementable in a machine and friendly enough to be understandable by a wide audience. In this paper, we propose a general approach to Explainable Artificial Intelligence in the case of neural architectures, showing how a mindful design of the networks leads to a family of interpretable deep learning models called Logic Explained Networks (LENs). LENs only require their inputs to be human-understandable predicates, and they provide explanations in terms of simple First-Order Logic (FOL) formulas involving such predicates. LENs are general enough to cover a large number of scenarios. Amongst them, we consider the case in which LENs are directly used as special classifiers with the capability of being explainable, or when they act as additional networks with the role of creating the conditions for making a black-box classifier explainable by FOL formulas. Despite supervised learning problems are mostly emphasized, we also show that LENs can learn and provide explanations in unsupervised learning settings. Experimental results on several datasets and tasks show that LENs may yield better classifications than established white-box models, such as decision trees and Bayesian rule lists, while providing more compact and meaningful explanations.
The need of predictive maintenance comes with an increasing number of incidents reported by monitoring systems and equipment/software users. In the front line, on-call engineers (OCEs) have to quickly assess the degree of severity of an incident and decide which service to contact for corrective actions. To automate these decisions, several predictive models have been proposed, but the most efficient models are opaque (say, black box), strongly limiting their adoption. In this paper, we propose an efficient black box model based on 170K incidents reported to our company over the last 7 years and emphasize on the need of automating triage when incidents are massively reported on thousands of servers running our product, an ERP. Recent developments in eXplainable Artificial Intelligence (XAI) help in providing global explanations to the model, but also, and most importantly, with local explanations for each model prediction/outcome. Sadly, providing a human with an explanation for each outcome is not conceivable when dealing with an important number of daily predictions. To address this problem, we propose an original data-mining method rooted in Subgroup Discovery, a pattern mining technique with the natural ability to group objects that share similar explanations of their black box predictions and provide a description for each group. We evaluate this approach and present our preliminary results which give us good hope towards an effective OCE's adoption. We believe that this approach provides a new way to address the problem of model agnostic outcome explanation.
Traffic accident anticipation is a vital function of Automated Driving Systems (ADSs) for providing a safety-guaranteed driving experience. An accident anticipation model aims to predict accidents promptly and accurately before they occur. Existing Artificial Intelligence (AI) models of accident anticipation lack a human-interpretable explanation of their decision-making. Although these models perform well, they remain a black-box to the ADS users, thus difficult to get their trust. To this end, this paper presents a Gated Recurrent Unit (GRU) network that learns spatio-temporal relational features for the early anticipation of traffic accidents from dashcam video data. A post-hoc attention mechanism named Grad-CAM is integrated into the network to generate saliency maps as the visual explanation of the accident anticipation decision. An eye tracker captures human eye fixation points for generating human attention maps. The explainability of network-generated saliency maps is evaluated in comparison to human attention maps. Qualitative and quantitative results on a public crash dataset confirm that the proposed explainable network can anticipate an accident on average 4.57 seconds before it occurs, with 94.02% average precision. In further, various post-hoc attention-based XAI methods are evaluated and compared. It confirms that the Grad-CAM chosen by this study can generate high-quality, human-interpretable saliency maps (with 1.23 Normalized Scanpath Saliency) for explaining the crash anticipation decision. Importantly, results confirm that the proposed AI model, with a human-inspired design, can outperform humans in the accident anticipation.
Advances in machine learning have led to graph neural network-based methods for drug discovery, yielding promising results in molecular design, chemical synthesis planning, and molecular property prediction. However, current graph neural networks (GNNs) remain of limited acceptance in drug discovery is limited due to their lack of interpretability. Although this major weakness has been mitigated by the development of explainable artificial intelligence (XAI) techniques, the "ground truth" assignment in most explainable tasks ultimately rests with subjective judgments by humans so that the quality of model interpretation is hard to evaluate in quantity. In this work, we first build three levels of benchmark datasets to quantitatively assess the interpretability of the state-of-the-art GNN models. Then we implemented recent XAI methods in combination with different GNN algorithms to highlight the benefits, limitations, and future opportunities for drug discovery. As a result, GradInput and IG generally provide the best model interpretability for GNNs, especially when combined with GraphNet and CMPNN. The integrated and developed XAI package is fully open-sourced and can be used by practitioners to train new models on other drug discovery tasks.
Machine learning solutions for pattern classification problems are nowadays widely deployed in society and industry. However, the lack of transparency and accountability of most accurate models often hinders their safe use. Thus, there is a clear need for developing explainable artificial intelligence mechanisms. There exist model-agnostic methods that summarize feature contributions, but their interpretability is limited to predictions made by black-box models. An open challenge is to develop models that have intrinsic interpretability and produce their own explanations, even for classes of models that are traditionally considered black boxes like (recurrent) neural networks. In this paper, we propose a Long-Term Cognitive Network for interpretable pattern classification of structured data. Our method brings its own mechanism for providing explanations by quantifying the relevance of each feature in the decision process. For supporting the interpretability without affecting the performance, the model incorporates more flexibility through a quasi-nonlinear reasoning rule that allows controlling nonlinearity. Besides, we propose a recurrence-aware decision model that evades the issues posed by the unique fixed point while introducing a deterministic learning algorithm to compute the tunable parameters. The simulations show that our interpretable model obtains competitive results when compared to state-of-the-art white and black-box models.
Strategies based on Explainable Artificial Intelligence - XAI have emerged in computing to promote a better understanding of predictions made by black box models. Most XAI measures used today explain these types of models, generating attribute rankings aimed at explaining the model, that is, the analysis of Attribute Importance of Model. There is no consensus on which XAI measure generates an overall explainability rank. For this reason, several proposals for tools have emerged (Ciu, Dalex, Eli5, Lofo, Shap and Skater). An experimental benchmark of explainable AI techniques capable of producing global explainability ranks based on tabular data related to different problems and ensemble models are presented herein. Seeking to answer questions such as "Are the explanations generated by the different measures the same, similar or different?" and "How does data complexity play along model explainability?" The results from the construction of 82 computational models and 592 ranks shed some light on the other side of the problem of explainability: dataset complexity!
Deep Neural Networks (DNNs) are known to be strong predictors, but their prediction strategies can rarely be understood. With recent advances in Explainable Artificial Intelligence (XAI), approaches are available to explore the reasoning behind those complex models' predictions. Among post-hoc attribution methods, Layer-wise Relevance Propagation (LRP) shows high performance. For deeper quantitative analysis, manual approaches exist, but without the right tools they are unnecessarily labor intensive. In this software paper, we introduce three software packages targeted at scientists to explore model reasoning using attribution approaches and beyond: (1) Zennit - a highly customizable and intuitive attribution framework implementing LRP and related approaches in PyTorch, (2) CoRelAy - a framework to easily and quickly construct quantitative analysis pipelines for dataset-wide analyses of explanations, and (3) ViRelAy - a web-application to interactively explore data, attributions, and analysis results. With this, we provide a standardized implementation solution for XAI, to contribute towards more reproducibility in our field.
The black-box nature of machine learning models limits their use in case-critical applications, raising faithful and ethical concerns that lead to trust crises. One possible way to mitigate this issue is to understand how a (mispredicted) decision is carved out from the decision boundary. This paper presents a human-in-the-loop approach to explain machine learning models using verbatim neighborhood manifestation. Contrary to most of the current eXplainable Artificial Intelligence (XAI) systems, which provide hit-or-miss approximate explanations, our approach generates the local decision boundary of the given instance and enables human intelligence to conclude the model behavior. Our method can be divided into three stages: 1) a neighborhood generation stage, which generates instances based on the given sample; 2) a classification stage, which yields classifications on the generated instances to carve out the local decision boundary and delineate the model behavior; and 3) a human-in-the-loop stage, which involves human to refine and explore the neighborhood of interest. In the generation stage, a generative model is used to generate the plausible synthetic neighbors around the given instance. After the classification stage, the classified neighbor instances provide a multifaceted understanding of the model behavior. Three intervention points are provided in the human-in-the-loop stage, enabling humans to leverage their own intelligence to interpret the model behavior. Several experiments on two datasets are conducted, and the experimental results demonstrate the potential of our proposed approach for boosting human understanding of the complex machine learning model.
A precise understanding of why units in an artificial network respond to certain stimuli would constitute a big step towards explainable artificial intelligence. One widely used approach towards this goal is to visualize unit responses via activation maximization. These synthetic feature visualizations are purported to provide humans with precise information about the image features that cause a unit to be activated - an advantage over other alternatives like strongly activating natural dataset samples. If humans indeed gain causal insight from visualizations, this should enable them to predict the effect of an intervention, such as how occluding a certain patch of the image (say, a dog's head) changes a unit's activation. Here, we test this hypothesis by asking humans to decide which of two square occlusions causes a larger change to a unit's activation. Both a large-scale crowdsourced experiment and measurements with experts show that on average the extremely activating feature visualizations by Olah et al. (2017) indeed help humans on this task ($68 \pm 4$% accuracy; baseline performance without any visualizations is $60 \pm 3$%). However, they do not provide any substantial advantage over other visualizations (such as e.g. dataset samples), which yield similar performance ($66\pm3$% to $67 \pm3$% accuracy). Taken together, we propose an objective psychophysical task to quantify the benefit of unit-level interpretability methods for humans, and find no evidence that a widely-used feature visualization method provides humans with better "causal understanding" of unit activations than simple alternative visualizations.
Explaining the predictions of opaque machine learning algorithms is an important and challenging task, especially as complex models are increasingly used to assist in high-stakes decisions such as those arising in healthcare and finance. Most popular tools for post-hoc explainable artificial intelligence (XAI) are either insensitive to context (e.g., feature attributions) or difficult to summarize (e.g., counterfactuals). In this paper, I introduce $\textit{rational Shapley values}$, a novel XAI method that synthesizes and extends these seemingly incompatible approaches in a rigorous, flexible manner. I leverage tools from decision theory and causal modeling to formalize and implement a pragmatic approach that resolves a number of known challenges in XAI. By pairing the distribution of random variables with the appropriate reference class for a given explanation task, I illustrate through theory and experiments how user goals and knowledge can inform and constrain the solution set in an iterative fashion. The method compares favorably to state of the art XAI tools in a range of quantitative and qualitative comparisons.
Deterministic frequency deviations (DFDs) critically affect power grid frequency quality and power system stability. A better understanding of these events is urgently needed as frequency deviations have been growing in the European grid in recent years. DFDs are partially explained by the rapid adjustment of power generation following the intervals of electricity trading, but this intuitive picture fails especially before and around noonday. In this article, we provide a detailed analysis of DFDs and their relation to external features using methods from explainable Artificial Intelligence. We establish a machine learning model that well describes the daily cycle of DFDs and elucidate key interdependencies using SHapley Additive exPlanations (SHAP). Thereby, we identify solar ramps as critical to explain patterns in the Rate of Change of Frequency (RoCoF).
The field of Explainable Artificial Intelligence (XAI) aims to build explainable and interpretable machine learning (or deep learning) methods without sacrificing prediction performance. Convolutional Neural Networks (CNNs) have been successful in making predictions, especially in image classification. However, these famous deep learning models use tens of millions of parameters based on a large number of pre-trained filters which have been repurposed from previous data sets. We propose a novel Interaction-based Convolutional Neural Network (ICNN) that does not make assumptions about the relevance of local information. Instead, we use a model-free Influence Score (I-score) to directly extract the influential information from images to form important variable modules. We demonstrate that the proposed method produces state-of-the-art prediction performance of 99.8% on a real-world data set classifying COVID-19 Chest X-ray images without sacrificing the explanatory power of the model. This proposed design can efficiently screen COVID-19 patients before human diagnosis, and will be the benchmark for addressing future XAI problems in large-scale data sets.
Explainable artificial intelligence has rapidly emerged since lawmakers have started requiring interpretable models for safety-critical domains. Concept-based neural networks have arisen as explainable-by-design methods as they leverage human-understandable symbols (i.e. concepts) to predict class memberships. However, most of these approaches focus on the identification of the most relevant concepts but do not provide concise, formal explanations of how such concepts are leveraged by the classifier to make predictions. In this paper, we propose a novel end-to-end differentiable approach enabling the extraction of logic explanations from neural networks using the formalism of First-Order Logic. The method relies on an entropy-based criterion which automatically identifies the most relevant concepts. We consider four different case studies to demonstrate that: (i) this entropy-based criterion enables the distillation of concise logic explanations in safety-critical domains from clinical data to computer vision; (ii) the proposed approach outperforms state-of-the-art white-box models in terms of classification accuracy and matches black box performances.
The main objective of eXplainable Artificial Intelligence (XAI) is to provide effective explanations for black-box classifiers. The existing literature lists many desirable properties for explanations to be useful, but there is no consensus on how to quantitatively evaluate explanations in practice. Moreover, explanations are typically used only to inspect black-box models, and the proactive use of explanations as a decision support is generally overlooked. Among the many approaches to XAI, a widely adopted paradigm is Local Linear Explanations - with LIME and SHAP emerging as state-of-the-art methods. We show that these methods are plagued by many defects including unstable explanations, divergence of actual implementations from the promised theoretical properties, and explanations for the wrong label. This highlights the need to have standard and unbiased evaluation procedures for Local Linear Explanations in the XAI field. In this paper we address the problem of identifying a clear and unambiguous set of metrics for the evaluation of Local Linear Explanations. This set includes both existing and novel metrics defined specifically for this class of explanations. All metrics have been included in an open Python framework, named LEAF. The purpose of LEAF is to provide a reference for end users to evaluate explanations in a standardised and unbiased way, and to guide researchers towards developing improved explainable techniques.
Explainable artificial intelligence is the attempt to elucidate the workings of systems too complex to be directly accessible to human cognition through suitable side-information referred to as "explanations". We present a trainable explanation module for convolutional image classifiers we call bounded logit attention (BLA). The BLA module learns to select a subset of the convolutional feature map for each input instance, which then serves as an explanation for the classifier's prediction. BLA overcomes several limitations of the instancewise feature selection method "learning to explain" (L2X) introduced by Chen et al. (2018): 1) BLA scales to real-world sized image classification problems, and 2) BLA offers a canonical way to learn explanations of variable size. Due to its modularity BLA lends itself to transfer learning setups and can also be employed as a post-hoc add-on to trained classifiers. Beyond explainability, BLA may serve as a general purpose method for differentiable approximation of subset selection. In a user study we find that BLA explanations are preferred over explanations generated by the popular (Grad-)CAM method.