Research Papers

Model Interpretation and Explainability: Towards Creating Transparency in Prediction Models

Donald Kridel, Jacob Dineen, Daniel Dolk +1

Explainable AI (XAI) has a counterpart in analytical modeling which we refer to as model explainability. We tackle the issue of model explainability in the context of prediction models. We analyze a dataset of loans from a credit card company and apply three stages: execute and compare four different prediction methods, apply the best known explainability techniques in the current literature to the model training sets to identify feature importance (FI) (static case), and finally to cross-check whether the FI set holds up under what if prediction scenarios for continuous and categorical variables (dynamic case). We found inconsistency in FI identification between the static and dynamic cases. We summarize the state of the art in model explainability and suggest further research to advance the field.

Despite the excelling performance of machine learning models, understanding their decisions remains a long-standing goal. Although commonly used attribution methods from explainable AI attempt to address this issue, they typically rely on associational rather than causal relationships. In this study, within the context of time series classification, we introduce a novel model-agnostic attribution method to assess the causal effect of concepts i.e., predefined segments within a time series, on classification outcomes. Our approach compares these causal attributions with closely related associational attributions, both theoretically and empirically. To estimate counterfactual outcomes, we use state-of-the-art diffusion models backed by state space models. We demonstrate the insights gained by our approach for a diverse set of qualitatively different time series classification tasks. Although causal and associational attributions might often share some similarities, in all cases they differ in important details, underscoring the risks associated with drawing causal conclusions from associational data alone. We believe that the proposed approach is also widely applicable in other domains to shed some light on the limits of associational attributions.

In this paper, we explore the feasibility of using a transformer-based, spatiotemporal attention network (STAN) for gradient-based time-series explanations. First, we trained the STAN model for video classifications using the global and local views of data and weakly supervised labels on time-series data (i.e. the type of an activity). We then leveraged a gradient-based XAI technique (e.g. saliency map) to identify salient frames of time-series data. According to the experiments using the datasets of four medically relevant activities, the STAN model demonstrated its potential to identify important frames of videos.

Mounting evidence in explainability for artificial intelligence (XAI) research suggests that good explanations should be tailored to individual tasks and should relate to concepts relevant to the task. However, building task specific explanations is time consuming and requires domain expertise which can be difficult to integrate into generic XAI methods. A promising approach towards designing useful task specific explanations with domain experts is based on compositionality of semantic concepts. Here, we present a novel approach that enables domain experts to quickly create concept-based explanations for computer vision tasks intuitively via natural language. Leveraging recent progress in deep generative methods we propose to generate visual concept-based prototypes via text-to-image methods. These prototypes are then used to explain predictions of computer vision models via a simple k-Nearest-Neighbors routine. The modular design of CoProNN is simple to implement, it is straightforward to adapt to novel tasks and allows for replacing the classification and text-to-image models as more powerful models are released. The approach can be evaluated offline against the ground-truth of predefined prototypes that can be easily communicated also to domain experts as they are based on visual concepts. We show that our strategy competes very well with other concept-based XAI approaches on coarse grained image classification tasks and may even outperform those methods on more demanding fine grained tasks. We demonstrate the effectiveness of our method for human-machine collaboration settings in qualitative and quantitative user studies. All code and experimental data can be found in our GitHub $\href{https://github.com/TeodorChiaburu/beexplainable}{repository}$.

Given the sheer volume of surgical procedures and the significant rate of postoperative fatalities, assessing and managing surgical complications has become a critical public health concern. Existing artificial intelligence (AI) tools for risk surveillance and diagnosis often lack adequate interpretability, fairness, and reproducibility. To address this, we proposed an Explainable AI (XAI) framework designed to answer five critical questions: why, why not, how, what if, and what else, with the goal of enhancing the explainability and transparency of AI models. We incorporated various techniques such as Local Interpretable Model-agnostic Explanations (LIME), SHapley Additive exPlanations (SHAP), counterfactual explanations, model cards, an interactive feature manipulation interface, and the identification of similar patients to address these questions. We showcased an XAI interface prototype that adheres to this framework for predicting major postoperative complications. This initial implementation has provided valuable insights into the vast explanatory potential of our XAI framework and represents an initial step towards its clinical adoption.

Deep learning models show significant potential for advancing AI-assisted medical diagnostics, particularly in detecting lung cancer through medical image modalities such as chest X-rays. However, the black-box nature of these models poses challenges to their interpretability and trustworthiness, limiting their adoption in clinical practice. This study examines both the interpretability and robustness of a high-performing lung cancer detection model based on InceptionV3, utilizing a public dataset of chest X-rays and radiological reports. We evaluate the clinical utility of multiple explainable AI (XAI) techniques, including both post-hoc and ante-hoc approaches, and find that existing methods often fail to provide clinically relevant explanations, displaying inconsistencies and divergence from expert radiologist assessments. To address these limitations, we collaborated with a radiologist to define diagnosis-specific clinical concepts and developed ClinicXAI, an expert-driven approach leveraging the concept bottleneck methodology. ClinicXAI generated clinically meaningful explanations which closely aligned with the practical requirements of clinicians while maintaining high diagnostic accuracy. We also assess the robustness of ClinicXAI in comparison to the original InceptionV3 model by subjecting both to a series of widely utilized adversarial attacks. Our analysis demonstrates that ClinicXAI exhibits significantly greater resilience to adversarial perturbations. These findings underscore the importance of incorporating domain expertise into the design of interpretable and robust AI systems for medical diagnostics, paving the way for more trustworthy and effective AI solutions in healthcare.

There is growing interest in AI systems that support human decision-making in high-stakes domains (e.g., medical diagnosis) to improve decision quality and reduce cognitive load. Mainstream approaches pair human experts with a machine-learning model, offloading low-risk decisions to the model so that experts can focus on cases that require their judgment. This separation of responsibilities setup, however, is inadequate for high-stakes scenarios. The expert may end up over-relying on the machine's decisions due to anchoring bias, thus losing the human oversight that is increasingly being required by regulatory agencies to ensure trustworthy AI. On the other hand, the expert is left entirely unassisted on the (typically hardest) decisions on which the model abstained. As a remedy, we introduce learning to guide (LTG), an alternative framework in which -- rather than taking control from the human expert -- the machine provides guidance useful for decision making, and the human is entirely responsible for coming up with a decision. In order to ensure guidance is interpretable and task-specific, we develop SLOG, an approach for turning any vision-language model into a capable generator of textual guidance by leveraging a modicum of human feedback. Our empirical evaluation highlights the promise of SLOG on both on a synthetic dataset and a challenging, real-world medical diagnosis task.

The field of explainable artificial intelligence (XAI) attempts to develop methods that provide insight into how complicated machine learning methods make predictions. Many methods of explanation have focused on the concept of feature attribution, a decomposition of the model's prediction into individual contributions corresponding to each input feature. In this work, we explore the problem of feature attribution in the context of Gaussian process regression (GPR). We take a principled approach to defining attributions under model uncertainty, extending the existing literature. We show that although GPR is a highly flexible and non-parametric approach, we can derive interpretable, closed-form expressions for the feature attributions. When using integrated gradients as an attribution method, we show that the attributions of a GPR model also follow a Gaussian process distribution, which quantifies the uncertainty in attribution arising from uncertainty in the model. We demonstrate, both through theory and experimentation, the versatility and robustness of this approach. We also show that, when applicable, the exact expressions for GPR attributions are both more accurate and less computationally expensive than the approximations currently used in practice. The source code for this project is freely available under MIT license at https://github.com/KurtButler/2024_attributions_paper.

In recent years, machine learning models, especially deep neural networks, have been widely used for classification tasks in the security domain. However, these models have been shown to be vulnerable to adversarial manipulation: small changes learned by an adversarial attack model, when applied to the input, can cause significant changes in the output. Most research on adversarial attacks and corresponding defense methods focuses only on scenarios where adversarial samples are directly generated by the attack model. In this study, we explore a more practical scenario in behavior-based authentication, where adversarial samples are collected from the attacker. The generated adversarial samples from the model are replicated by attackers with a certain level of discrepancy. We propose an eXplainable AI (XAI) based defense strategy against adversarial attacks in such scenarios. A feature selector, trained with our method, can be used as a filter in front of the original authenticator. It filters out features that are more vulnerable to adversarial attacks or irrelevant to authentication, while retaining features that are more robust. Through comprehensive experiments, we demonstrate that our XAI based defense strategy is effective against adversarial attacks and outperforms other defense strategies, such as adversarial training and defensive distillation.

Recent technological advancements in artificial intelligence and computer vision have enabled gait analysis on portable devices such as cell phones. However, most state-of-the-art vision-based systems still impose numerous constraints for capturing a patient's video, such as using a static camera and maintaining a specific distance from it. While these constraints are manageable under professional observation, they pose challenges in home settings. Another issue with most vision-based systems is their output, typically a classification label and confidence value, whose reliability is often questioned by medical professionals. This paper addresses these challenges by presenting a novel system for gait analysis robust to camera movements and providing explanations for its output. The study utilizes a dataset comprising videos of subjects wearing two types of Knee Ankle Foot Orthosis (KAFO), namely "Locked Knee" and "Semi-flexion," for mobility, along with metadata and ground truth for explanations. The ground truth highlights the statistical significance of seven features captured using motion capture systems to differentiate between the two gaits. To address camera movement challenges, the proposed system employs super-resolution and pose estimation during pre-processing. It then identifies the seven features - Stride Length, Step Length and Duration of single support of orthotic and non-orthotic leg, Cadence, and Speed - using the skeletal output of pose estimation. These features train a multi-layer perceptron, with its output explained by highlighting the features' contribution to classification. While most state-of-the-art systems struggle with processing the video or training on the proposed dataset, our system achieves an average accuracy of 94%. The model's explainability is validated using ground truth and can be considered reliable.

Investigating the Impact of Model Instability on Explanations and Uncertainty

Sara Vera Marjanović, Isabelle Augenstein, Christina Lioma

Explainable AI methods facilitate the understanding of model behaviour, yet, small, imperceptible perturbations to inputs can vastly distort explanations. As these explanations are typically evaluated holistically, before model deployment, it is difficult to assess when a particular explanation is trustworthy. Some studies have tried to create confidence estimators for explanations, but none have investigated an existing link between uncertainty and explanation quality. We artificially simulate epistemic uncertainty in text input by introducing noise at inference time. In this large-scale empirical study, we insert different levels of noise perturbations and measure the effect on the output of pre-trained language models and different uncertainty metrics. Realistic perturbations have minimal effect on performance and explanations, yet masking has a drastic effect. We find that high uncertainty doesn't necessarily imply low explanation plausibility; the correlation between the two metrics can be moderately positive when noise is exposed during the training process. This suggests that noise-augmented models may be better at identifying salient tokens when uncertain. Furthermore, when predictive and epistemic uncertainty measures are over-confident, the robustness of a saliency map to perturbation can indicate model stability issues. Integrated Gradients shows the overall greatest robustness to perturbation, while still showing model-specific patterns in performance; however, this phenomenon is limited to smaller Transformer-based language models.

Recently, there has been a surge of explainable AI (XAI) methods driven by the need for understanding machine learning model behaviors in high-stakes scenarios. However, properly evaluating the effectiveness of the XAI methods inevitably requires the involvement of human subjects, and conducting human-centered benchmarks is challenging in a number of ways: designing and implementing user studies is complex; numerous design choices in the design space of user study lead to problems of reproducibility; and running user studies can be challenging and even daunting for machine learning researchers. To address these challenges, this paper presents OpenHEXAI, an open-source framework for human-centered evaluation of XAI methods. OpenHEXAI features (1) a collection of diverse benchmark datasets, pre-trained models, and post hoc explanation methods; (2) an easy-to-use web application for user study; (3) comprehensive evaluation metrics for the effectiveness of post hoc explanation methods in the context of human-AI decision making tasks; (4) best practice recommendations of experiment documentation; and (5) convenient tools for power analysis and cost estimation. OpenHEAXI is the first large-scale infrastructural effort to facilitate human-centered benchmarks of XAI methods. It simplifies the design and implementation of user studies for XAI methods, thus allowing researchers and practitioners to focus on the scientific questions. Additionally, it enhances reproducibility through standardized designs. Based on OpenHEXAI, we further conduct a systematic benchmark of four state-of-the-art post hoc explanation methods and compare their impacts on human-AI decision making tasks in terms of accuracy, fairness, as well as users' trust and understanding of the machine learning model.