We present a simple yet highly generalizable method for explaining interacting parts within a neural network's reasoning process. First, we design an algorithm based on cross derivatives for computing statistical interaction effects between individual features, which is generalized to both 2-way and higher-order (3-way or more) interactions. We present results side by side with a weight-based attribution technique, corroborating that cross derivatives are a superior metric for both 2-way and higher-order interaction detection. Moreover, we extend the use of cross derivatives as an explanatory device in neural networks to the computer vision setting by expanding Grad-CAM, a popular gradient-based explanatory tool for CNNs, to the higher order. While Grad-CAM can only explain the importance of individual objects in images, our method, which we call Taylor-CAM, can explain a neural network's relational reasoning across multiple objects. We show the success of our explanations both qualitatively and quantitatively, including with a user study. We will release all code as a tool package to facilitate explainable deep learning.
The significant advances in autonomous systems together with an immensely wider application domain have increased the need for trustable intelligent systems. Explainable artificial intelligence is gaining considerable attention among researchers and developers to address this requirement. Although there is an increasing number of works on interpretable and transparent machine learning algorithms, they are mostly intended for the technical users. Explanations for the end-user have been neglected in many usable and practical applications. In this work, we present the Contextual Importance (CI) and Contextual Utility (CU) concepts to extract explanations that are easily understandable by experts as well as novice users. This method explains the prediction results without transforming the model into an interpretable one. We present an example of providing explanations for linear and non-linear models to demonstrate the generalizability of the method. CI and CU are numerical values that can be represented to the user in visuals and natural language form to justify actions and explain reasoning for individual instances, situations, and contexts. We show the utility of explanations in car selection example and Iris flower classification by presenting complete (i.e. the causes of an individual prediction) and contrastive explanation (i.e. contrasting instance against the instance of interest). The experimental results show the feasibility and validity of the provided explanation methods.
Explainable Artificial Intelligence (XAI) has experienced a significant growth over the last few years. This is due to the widespread application of machine learning, particularly deep learning, that has led to the development of highly accurate models but lack explainability and interpretability. A plethora of methods to tackle this problem have been proposed, developed and tested. This systematic review contributes to the body of knowledge by clustering these methods with a hierarchical classification system with four main clusters: review articles, theories and notions, methods and their evaluation. It also summarises the state-of-the-art in XAI and recommends future research directions.
Traditionally, abnormal heart sound classification is framed as a three-stage process. The first stage involves segmenting the phonocardiogram to detect fundamental heart sounds; after which features are extracted and classification is performed. Some researchers in the field argue the segmentation step is an unwanted computational burden, whereas others embrace it as a prior step to feature extraction. When comparing accuracies achieved by studies that have segmented heart sounds before analysis with those who have overlooked that step, the question of whether to segment heart sounds before feature extraction is still open. In this study, we explicitly examine the importance of heart sound segmentation as a prior step for heart sound classification, and then seek to apply the obtained insights to propose a robust classifier for abnormal heart sound detection. Furthermore, recognizing the pressing need for explainable Artificial Intelligence (AI) models in the medical domain, we also unveil hidden representations learned by the classifier using model interpretation techniques. Experimental results demonstrate that the segmentation plays an essential role in abnormal heart sound classification. Our new classifier is also shown to be robust, stable and most importantly, explainable, with an accuracy of almost 100% on the widely used PhysioNet dataset.
Explainable Artificial Intelligence (or xAI) has become an important research topic in the fields of Machine Learning and Deep Learning. In this paper, we propose a Genetic Programming (GP) based approach, named Genetic Programming Explainer (GPX), to the problem of explaining decisions computed by AI systems. The method generates a noise set located in the neighborhood of the point of interest, whose prediction should be explained, and fits a local explanation model for the analyzed sample. The tree structure generated by GPX provides a comprehensible analytical, possibly non-linear, symbolic expression which reflects the local behavior of the complex model. We considered three machine learning techniques that can be recognized as complex black-box models: Random Forest, Deep Neural Network and Support Vector Machine in twenty data sets for regression and classifications problems. Our results indicate that the GPX is able to produce more accurate understanding of complex models than the state of the art. The results validate the proposed approach as a novel way to deploy GP to improve interpretability.
A novel evolutionary approach for Explainable Artificial Intelligence is presented: the "Evolved Explanations" model (EvEx). This methodology consists in combining Local Interpretable Model Agnostic Explanations (LIME) with Multi-Objective Genetic Algorithms to allow for automated segmentation parameter tuning in image classification tasks. In this case, the dataset studied is Patch-Camelyon, comprised of patches from pathology whole slide images. A publicly available Convolutional Neural Network (CNN) was trained on this dataset to provide a binary classification for presence/absence of lymph node metastatic tissue. In turn, the classifications are explained by means of evolving segmentations, seeking to optimize three evaluation goals simultaneously. The final explanation is computed as the mean of all explanations generated by Pareto front individuals, evolved by the developed genetic algorithm. To enhance reproducibility and traceability of the explanations, each of them was generated from several different seeds, randomly chosen. The observed results show remarkable agreement between different seeds. Despite the stochastic nature of LIME explanations, regions of high explanation weights proved to have good agreement in the heat maps, as computed by pixel-wise relative standard deviations. The found heat maps coincide with expert medical segmentations, which demonstrates that this methodology can find high quality explanations (according to the evaluation metrics), with the novel advantage of automated parameter fine tuning. These results give additional insight into the inner workings of neural network black box decision making for medical data.
Explainable Artificial Intelligence (XAI), i.e., the development of more transparent and interpretable AI models, has gained increased traction over the last few years. This is due to the fact that, in conjunction with their growth into powerful and ubiquitous tools, AI models exhibit one detrimential characteristic: a performance-transparency trade-off. This describes the fact that the more complex a model's inner workings, the less clear it is how its predictions or decisions were achieved. But, especially considering Machine Learning (ML) methods like Reinforcement Learning (RL) where the system learns autonomously, the necessity to understand the underlying reasoning for their decisions becomes apparent. Since, to the best of our knowledge, there exists no single work offering an overview of Explainable Reinforcement Learning (XRL) methods, this survey attempts to address this gap. We give a short summary of the problem, a definition of important terms, and offer a classification and assessment of current XRL methods. We found that a) the majority of XRL methods function by mimicking and simplifying a complex model instead of designing an inherently simple one, and b) XRL (and XAI) methods often neglect to consider the human side of the equation, not taking into account research from related fields like psychology or philosophy. Thus, an interdisciplinary effort is needed to adapt the generated explanations to a (non-expert) human user in order to effectively progress in the field of XRL and XAI in general.
Black-box Artificial Intelligence (AI) methods, e.g. deep neural networks, have been widely utilized to build predictive models that can extract complex relationships in a dataset and make predictions for new unseen data records. However, it is difficult to trust decisions made by such methods since their inner working and decision logic is hidden from the user. Explainable Artificial Intelligence (XAI) refers to systems that try to explain how a black-box AI model produces its outcomes. Post-hoc XAI methods approximate the behavior of a black-box by extracting relationships between feature values and the predictions. Perturbation-based and decision set methods are among commonly used post-hoc XAI systems. The former explanators rely on random perturbations of data records to build local or global linear models that explain individual predictions or the whole model. The latter explanators use those feature values that appear more frequently to construct a set of decision rules that produces the same outcomes as the target black-box. However, these two classes of XAI methods have some limitations. Random perturbations do not take into account the distribution of feature values in different subspaces, leading to misleading approximations. Decision sets only pay attention to frequent feature values and miss many important correlations between features and class labels that appear less frequently but accurately represent decision boundaries of the model. In this paper, we address the above challenges by proposing an explanation method named Confident Itemsets Explanation (CIE). We introduce confident itemsets, a set of feature values that are highly correlated to a specific class label. CIE utilizes confident itemsets to discretize the whole decision space of a model to smaller subspaces.
The rapid growth of research in explainable artificial intelligence (XAI) follows on two substantial developments. First, the enormous application success of modern machine learning methods, especially deep and reinforcement learning, which have created high expectations for industrial, commercial and social value. Second, the emergence of concern for creating trusted AI systems, including the creation of regulatory principles to ensure transparency and trust of AI systems.These two threads have created a kind of "perfect storm" of research activity, all eager to create and deliver it any set of tools and techniques to address the XAI demand. As some surveys of current XAI suggest, there is yet to appear a principled framework that respects the literature of explainability in the history of science, and which provides a basis for the development of a framework for transparent XAI. Here we intend to provide a strategic inventory of XAI requirements, demonstrate their connection to a history of XAI ideas, and synthesize those ideas into a simple framework to calibrate five successive levels of XAI.
Explaining the behaviors of deep neural networks, usually considered as black boxes, is critical especially when they are now being adopted over diverse aspects of human life. Taking the advantages of interpretable machine learning (interpretable ML), this paper proposes a novel tool called Catastrophic Forgetting Dissector (or CFD) to explain catastrophic forgetting in continual learning settings. We also introduce a new method called Critical Freezing based on the observations of our tool. Experiments on ResNet articulate how catastrophic forgetting happens, particularly showing which components of this famous network are forgetting. Our new continual learning algorithm defeats various recent techniques by a significant margin, proving the capability of the investigation. Critical freezing not only attacks catastrophic forgetting but also exposes explainability.
Explainable artificial intelligence provides tools to better understand predictive models and their decisions, but many such methods are limited to producing insights with respect to a single class. When generating explanations for several classes, reasoning over them to obtain a complete view may be difficult since they can present competing or contradictory evidence. To address this challenge we introduce the novel paradigm of multi-class explanations. We outline the theory behind such techniques and propose a local surrogate model based on multi-output regression trees -- called LIMEtree -- that offers faithful and consistent explanations of multiple classes for individual predictions while being post-hoc, model-agnostic and data-universal. On top of strong fidelity guarantees, our implementation delivers a range of diverse explanation types, including counterfactual statements favoured in the literature. We evaluate our algorithm with respect to explainability desiderata, through quantitative experiments and via a pilot user study, on image and tabular data classification tasks, comparing it to LIME, which is a state-of-the-art surrogate explainer. Our contributions demonstrate the benefits of multi-class explanations and wide-ranging advantages of our method across a diverse set of scenarios.
The growing need for in-depth analysis of predictive models leads to a series of new methods for explaining their local and global properties. Which of these methods is the best? It turns out that this is an ill-posed question. One cannot sufficiently explain a black-box machine learning model using a single method that gives only one perspective. Isolated explanations are prone to misunderstanding, leading to wrong or simplistic reasoning. This problem is known as the Rashomon effect and refers to diverse, even contradictory, interpretations of the same phenomenon. Surprisingly, most methods developed for explainable and responsible machine learning focus on a single-aspect of the model behavior. In contrast, we showcase the problem of explainability as an interactive and sequential analysis of a model. This paper proposes how different Explanatory Model Analysis (EMA) methods complement each other and discusses why it is essential to juxtapose them. The introduced process of Interactive EMA (IEMA) derives from the algorithmic side of explainable machine learning and aims to embrace ideas developed in cognitive sciences. We formalize the grammar of IEMA to describe potential human-model dialogues. It is implemented in a widely used human-centered open-source software framework that adopts interactivity, customizability and automation as its main traits. We conduct a user study to evaluate the usefulness of IEMA, which indicates that an interactive sequential analysis of a model increases the performance and confidence of human decision making.
Recent applications of autonomous agents and robots, such as self-driving cars, scenario-based trainers, exploration robots, and service robots have brought attention to crucial trust-related challenges associated with the current generation of artificial intelligence (AI) systems. AI systems based on the connectionist deep learning neural network approach lack capabilities of explaining their decisions and actions to others, despite their great successes. Without symbolic interpretation capabilities, they are black boxes, which renders their decisions or actions opaque, making it difficult to trust them in safety-critical applications. The recent stance on the explainability of AI systems has witnessed several approaches on eXplainable Artificial Intelligence (XAI); however, most of the studies have focused on data-driven XAI systems applied in computational sciences. Studies addressing the increasingly pervasive goal-driven agents and robots are still missing. This paper reviews approaches on explainable goal-driven intelligent agents and robots, focusing on techniques for explaining and communicating agents perceptual functions (example, senses, and vision) and cognitive reasoning (example, beliefs, desires, intention, plans, and goals) with humans in the loop. The review highlights key strategies that emphasize transparency, understandability, and continual learning for explainability. Finally, the paper presents requirements for explainability and suggests a roadmap for the possible realization of effective goal-driven explainable agents and robots.
Interest in the field of Explainable Artificial Intelligence has been growing for decades and has accelerated recently. As Artificial Intelligence models have become more complex, and often more opaque, with the incorporation of complex machine learning techniques, explainability has become more critical. Recently, researchers have been investigating and tackling explainability with a user-centric focus, looking for explanations to consider trustworthiness, comprehensibility, explicit provenance, and context-awareness. In this chapter, we leverage our survey of explanation literature in Artificial Intelligence and closely related fields and use these past efforts to generate a set of explanation types that we feel reflect the expanded needs of explanation for today's artificial intelligence applications. We define each type and provide an example question that would motivate the need for this style of explanation. We believe this set of explanation types will help future system designers in their generation and prioritization of requirements and further help generate explanations that are better aligned to users' and situational needs.
Explainability has been an important goal since the early days of Artificial Intelligence. Several approaches for producing explanations have been developed. However, many of these approaches were tightly coupled with the capabilities of the artificial intelligence systems at the time. With the proliferation of AI-enabled systems in sometimes critical settings, there is a need for them to be explainable to end-users and decision-makers. We present a historical overview of explainable artificial intelligence systems, with a focus on knowledge-enabled systems, spanning the expert systems, cognitive assistants, semantic applications, and machine learning domains. Additionally, borrowing from the strengths of past approaches and identifying gaps needed to make explanations user- and context-focused, we propose new definitions for explanations and explainable knowledge-enabled systems.
Color and structure are the two pillars that construct an image. Usually, the structure is well expressed through a rich spectrum of colors, allowing objects in an image to be recognized by neural networks. However, under extreme limitations of color space, the structure tends to vanish, and thus a neural network might fail to understand the image. Interested in exploring this interplay between color and structure, we study the scientific problem of identifying and preserving the most informative image structures while constraining the color space to just a few bits, such that the resulting image can be recognized with possibly high accuracy. To this end, we propose a color quantization network, ColorCNN, which learns to structure the images from the classification loss in an end-to-end manner. Given a color space size, ColorCNN quantizes colors in the original image by generating a color index map and an RGB color palette. Then, this color-quantized image is fed to a pre-trained task network to evaluate its performance. In our experiment, with only a 1-bit color space (i.e., two colors), the proposed network achieves 82.1% top-1 accuracy on the CIFAR10 dataset, outperforming traditional color quantization methods by a large margin. For applications, when encoded with PNG, the proposed color quantization shows superiority over other image compression methods in the extremely low bit-rate regime. The code is available at: https://github.com/hou-yz/color_distillation.
Background / introduction. Vector symbolic architectures (VSA) are a viable approach for the hyperdimensional representation of symbolic data, such as documents, syntactic structures, or semantic frames. Methods. We present a rigorous mathematical framework for the representation of phrase structure trees and parse trees of context-free grammars (CFG) in Fock space, i.e. infinite-dimensional Hilbert space as being used in quantum field theory. We define a novel normal form for CFG by means of term algebras. Using a recently developed software toolbox, called FockBox, we construct Fock space representations for the trees built up by a CFG left-corner (LC) parser. Results. We prove a universal representation theorem for CFG term algebras in Fock space and illustrate our findings through a low-dimensional principal component projection of the LC parser states. Conclusions. Our approach could leverage the development of VSA for explainable artificial intelligence (XAI) by means of hyperdimensional deep neural computation. It could be of significance for the improvement of cognitive user interfaces and other applications of VSA in machine learning.
The application of unsupervised learning approaches, and in particular of clustering techniques, represents a powerful exploration means for the analysis of network measurements. Discovering underlying data characteristics, grouping similar measurements together, and identifying eventual patterns of interest are some of the applications which can be tackled through clustering. Being unsupervised, clustering does not always provide precise and clear insight into the produced output, especially when the input data structure and distribution are complex and difficult to grasp. In this paper we introduce EXPLAIN-IT, a methodology which deals with unlabeled data, creates meaningful clusters, and suggests an explanation to the clustering results for the end-user. EXPLAIN-IT relies on a novel explainable Artificial Intelligence (AI) approach, which allows to understand the reasons leading to a particular decision of a supervised learning-based model, additionally extending its application to the unsupervised learning domain. We apply EXPLAIN-IT to the problem of YouTube video quality classification under encrypted traffic scenarios, showing promising results.
In this paper I argue that the search for explainable models and interpretable decisions in AI must be reformulated in terms of the broader project of offering a pragmatic and naturalistic account of understanding in AI. Intuitively, the purpose of providing an explanation of a model or a decision is to make it understandable to its stakeholders. But without a previous grasp of what it means to say that an agent understands a model or a decision, the explanatory strategies will lack a well-defined goal. Aside from providing a clearer objective for XAI, focusing on understanding also allows us to relax the factivity condition on explanation, which is impossible to fulfill in many machine learning models, and to focus instead on the pragmatic conditions that determine the best fit between a model and the methods and devices deployed to understand it. After an examination of the different types of understanding discussed in the philosophical and psychological literature, I conclude that interpretative or approximation models not only provide the best way to achieve the objectual understanding of a machine learning model, but are also a necessary condition to achieve post-hoc interpretability. This conclusion is partly based on the shortcomings of the purely functionalist approach to post-hoc interpretability that seems to be predominant in most recent literature.
Recently we see a rising number of methods in the field of eXplainable Artificial Intelligence. To our surprise, their development is driven by model developers rather than a study of needs for human end users. The analysis of needs, if done, takes the form of an A/B test rather than a study of open questions. To answer the question "What would a human operator like to ask the ML model?" we propose a conversational system explaining decisions of the predictive model. In this experiment, we developed a chatbot called dr_ant to talk about machine learning model trained to predict survival odds on Titanic. People can talk with dr_ant about different aspects of the model to understand the rationale behind its predictions. Having collected a corpus of 1000+ dialogues, we analyse the most common types of questions that users would like to ask. To our knowledge, it is the first study which uses a conversational system to collect the needs of human operators from the interactive and iterative dialogue explorations of a predictive model.