We introduce a novel ensemble approach for feature selection based on hierarchical stacking for non-stationarity and/or a limited number of samples with a large number of features. Our approach exploits the co-dependency between features using a hierarchical structure. Initially, a machine learning model is trained using a subset of features, and then the output of the model is updated using other algorithms in a hierarchical manner with the remaining features to minimize the target loss. This hierarchical structure allows for flexible depth and feature selection. By exploiting feature co-dependency hierarchically, our proposed approach overcomes the limitations of traditional feature selection methods and feature importance scores. The effectiveness of the approach is demonstrated on synthetic and well-known real-life datasets, providing significant scalable and stable performance improvements compared to the traditional methods and the state-of-the-art approaches. We also provide the source code of our approach to facilitate further research and replicability of our results.
The United Nations has identified improving food security and reducing hunger as essential components of its sustainable development goals. As of 2021, approximately 828 million people worldwide are experiencing hunger and malnutrition, with numerous fatalities reported. Climate change significantly impacts agricultural land suitability, potentially leading to severe food shortages and subsequent social and political conflicts. To address this pressing issue, we have developed a machine learning-based approach to predict the risk of substantial land suitability degradation and changes in irrigation patterns. Our study focuses on Central Eurasia, a region burdened with economic and social challenges. This study represents a pioneering effort in utilizing machine learning methods to assess the impact of climate change on agricultural land suitability under various carbon emissions scenarios. Through comprehensive feature importance analysis, we unveil specific climate and terrain characteristics that exert influence on land suitability. Our approach achieves remarkable accuracy, offering policymakers invaluable insights to facilitate informed decisions aimed at averting a humanitarian crisis, including strategies such as the provision of additional water and fertilizers. This research underscores the tremendous potential of machine learning in addressing global challenges, with a particular emphasis on mitigating hunger and malnutrition.
Explainable artificial intelligence (XAI) is one of the most intensively developed area of AI in recent years. It is also one of the most fragmented with multiple methods that focus on different aspects of explanations. This makes difficult to obtain the full spectrum of explanation at once in a compact and consistent way. To address this issue, we present Local Universal Explainer (LUX), which is a rule-based explainer that can generate factual, counterfactual and visual explanations. It is based on a modified version of decision tree algorithms that allows for oblique splits and integration with feature importance XAI methods such as SHAP. It limits the use data generation in opposite to other algorithms, but is focused on selecting local concepts in a form of high-density clusters of real data that have the highest impact on forming the decision boundary of the explained model and generating artificial samples with novel SHAP-guided sampling algorithm. We tested our method on real and synthetic datasets and compared it with state-of-the-art rule-based explainers such as LORE, EXPLAN and Anchor. Our method outperforms the existing approaches in terms of simplicity, fidelity, representativeness, and consistency.
Unsupervised speech disentanglement aims at separating fast varying from slowly varying components of a speech signal. In this contribution, we take a closer look at the embedding vector representing the slowly varying signal components, commonly named the speaker embedding vector. We ask, which properties of a speaker's voice are captured and investigate to which extent do individual embedding vector components sign responsible for them, using the concept of Shapley values. Our findings show that certain speaker-specific acoustic-phonetic properties can be fairly well predicted from the speaker embedding, while the investigated more abstract voice quality features cannot.
In order to trust the predictions of a machine learning algorithm, it is necessary to understand the factors that contribute to those predictions. In the case of probabilistic and uncertainty-aware models, it is necessary to understand not only the reasons for the predictions themselves, but also the reasons for the model's level of confidence in those predictions. In this paper, we show how existing methods in explainability can be extended to uncertainty-aware models and how such extensions can be used to understand the sources of uncertainty in a model's predictive distribution. In particular, by adapting permutation feature importance, partial dependence plots, and individual conditional expectation plots, we demonstrate that novel insights into model behaviour may be obtained and that these methods can be used to measure the impact of features on both the entropy of the predictive distribution and the log-likelihood of the ground truth labels under that distribution. With experiments using both synthetic and real-world data, we demonstrate the utility of these approaches to understand both the sources of uncertainty and their impact on model performance.
Mental disorders impact the lives of millions of people globally, not only impeding their day-to-day lives but also markedly reducing life expectancy. This paper addresses the persistent challenge of predicting mortality in patients with mental diagnoses using predictive machine-learning models with electronic health records (EHR). Data from patients with mental disease diagnoses were extracted from the well-known clinical MIMIC-III data set utilizing demographic, prescription, and procedural information. Four machine learning algorithms (Logistic Regression, Random Forest, Support Vector Machine, and K-Nearest Neighbors) were used, with results indicating that Random Forest and Support Vector Machine models outperformed others, with AUC scores of 0.911. Feature importance analysis revealed that drug prescriptions, particularly Morphine Sulfate, play a pivotal role in prediction. We applied a variety of machine learning algorithms to predict 30-day mortality followed by feature importance analysis. This study can be used to assist hospital workers in identifying at-risk patients to reduce excess mortality.
The effectiveness of network intrusion detection systems, predominantly based on machine learning, are highly influenced by the dataset they are trained on. Ensuring an accurate reflection of the multifaceted nature of benign and malicious traffic in these datasets is essential for creating models capable of recognizing and responding to a wide array of intrusion patterns. However, existing datasets often fall short, lacking the necessary diversity and alignment with the contemporary network environment, thereby limiting the effectiveness of intrusion detection. This paper introduces TII-SSRC-23, a novel and comprehensive dataset designed to overcome these challenges. Comprising a diverse range of traffic types and subtypes, our dataset is a robust and versatile tool for the research community. Additionally, we conduct a feature importance analysis, providing vital insights into critical features for intrusion detection tasks. Through extensive experimentation, we also establish firm baselines for supervised and unsupervised intrusion detection methodologies using our dataset, further contributing to the advancement and adaptability of intrusion detection models in the rapidly changing landscape of network security. Our dataset is available at https://kaggle.com/datasets/daniaherzalla/tii-ssrc-23.
While deep neural networks have excellent results in many fields, they are susceptible to interference from attacking samples resulting in erroneous judgments. Feature-level attacks are one of the effective attack types, which targets the learnt features in the hidden layers to improve its transferability across different models. Yet it is observed that the transferability has been largely impacted by the neuron importance estimation results. In this paper, a double adversarial neuron attribution attack method, termed `DANAA', is proposed to obtain more accurate feature importance estimation. In our method, the model outputs are attributed to the middle layer based on an adversarial non-linear path. The goal is to measure the weight of individual neurons and retain the features that are more important towards transferability. We have conducted extensive experiments on the benchmark datasets to demonstrate the state-of-the-art performance of our method. Our code is available at: https://github.com/Davidjinzb/DANAA
This paper discusses predictive performance and processes undertaken on flight pricing data utilizing r2(r-square) and RMSE that leverages a large dataset, originally from Expedia.com, consisting of approximately 20 million records or 4.68 gigabytes. The project aims to determine the best models usable in the real world to predict airline ticket fares for non-stop flights across the US. Therefore, good generalization capability and optimized processing times are important measures for the model. We will discover key business insights utilizing feature importance and discuss the process and tools used for our analysis. Four regression machine learning algorithms were utilized: Random Forest, Gradient Boost Tree, Decision Tree, and Factorization Machines utilizing Cross Validator and Training Validator functions for assessing performance and generalization capability.
The 2022 National Defense Strategy of the United States listed climate change as a serious threat to national security. Climate intervention methods, such as stratospheric aerosol injection, have been proposed as mitigation strategies, but the downstream effects of such actions on a complex climate system are not well understood. The development of algorithmic techniques for quantifying relationships between source and impact variables related to a climate event (i.e., a climate pathway) would help inform policy decisions. Data-driven deep learning models have become powerful tools for modeling highly nonlinear relationships and may provide a route to characterize climate variable relationships. In this paper, we explore the use of an echo state network (ESN) for characterizing climate pathways. ESNs are a computationally efficient neural network variation designed for temporal data, and recent work proposes ESNs as a useful tool for forecasting spatio-temporal climate data. Like other neural networks, ESNs are non-interpretable black-box models, which poses a hurdle for understanding variable relationships. We address this issue by developing feature importance methods for ESNs in the context of spatio-temporal data to quantify variable relationships captured by the model. We conduct a simulation study to assess and compare the feature importance techniques, and we demonstrate the approach on reanalysis climate data. In the climate application, we select a time period that includes the 1991 volcanic eruption of Mount Pinatubo. This event was a significant stratospheric aerosol injection, which we use as a proxy for an artificial stratospheric aerosol injection. Using the proposed approach, we are able to characterize relationships between pathway variables associated with this event.
In this work, we instantiate a novel perturbation-based multi-class explanation framework, LIPEx (Locally Interpretable Probabilistic Explanation). We demonstrate that LIPEx not only locally replicates the probability distributions output by the widely used complex classification models but also provides insight into how every feature deemed to be important affects the prediction probability for each of the possible classes. We achieve this by defining the explanation as a matrix obtained via regression with respect to the Hellinger distance in the space of probability distributions. Ablation tests on text and image data, show that LIPEx-guided removal of important features from the data causes more change in predictions for the underlying model than similar tests based on other saliency-based or feature importance-based Explainable AI (XAI) methods. It is also shown that compared to LIME, LIPEx is more data efficient in terms of using a lesser number of perturbations of the data to obtain a reliable explanation. This data-efficiency is seen to manifest as LIPEx being able to compute its explanation matrix around 53% faster than all-class LIME, for classification experiments with text data.
Anomaly Detection (AD) focuses on identifying unusual behaviors in complex datasets. Machine Learning (ML) algorithms and Decision Support Systems (DSSs) provide effective solutions for AD, but detecting anomalies alone may not be enough, especially in engineering, where diagnostics and maintenance are crucial. Users need clear explanations to support root cause analysis and build trust in the model. The unsupervised nature of AD, however, makes interpretability a challenge. This paper introduces Extended Isolation Forest Feature Importance (ExIFFI), a method that explains predictions made by Extended Isolation Forest (EIF) models, which split data using hyperplanes. ExIFFI provides explanations at both global and local levels by leveraging feature importance. We also present an improved version, Enhanced Extended Isolation Forest (EIF+), designed to enhance the model's ability to detect unseen anomalies through a revised splitting strategy. Using five synthetic and eleven real-world datasets, we conduct a comparative analysis, evaluating unsupervised AD methods with the Average Precision metric. EIF+ consistently outperforms EIF across all datasets when trained without anomalies, demonstrating better generalization. To assess ExIFFI's interpretability, we introduce the Area Under the Curve of Feature Selection (AUC\_FS), a novel metric using feature selection as a proxy task. ExIFFI outperforms other unsupervised interpretability methods on 8 of 11 real-world datasets and successfully identifies anomalous features in synthetic datasets. When trained only on inliers, ExIFFI also outperforms competing models on real-world data and accurately detects anomalous features in synthetic datasets. We provide open-source code to encourage further research and reproducibility.
Transferring a pretrained model to a downstream task can be as easy as conducting linear probing with target data, that is, training a linear classifier upon frozen features extracted from the pretrained model. As there may exist significant gaps between pretraining and downstream datasets, one may ask whether all dimensions of the pretrained features are useful for a given downstream task. We show that, for linear probing, the pretrained features can be extremely redundant when the downstream data is scarce, or few-shot. For some cases such as 5-way 1-shot tasks, using only 1\% of the most important feature dimensions is able to recover the performance achieved by using the full representation. Interestingly, most dimensions are redundant only under few-shot settings and gradually become useful when the number of shots increases, suggesting that feature redundancy may be the key to characterizing the "few-shot" nature of few-shot transfer problems. We give a theoretical understanding of this phenomenon and show how dimensions with high variance and small distance between class centroids can serve as confounding factors that severely disturb classification results under few-shot settings. As an attempt at solving this problem, we find that the redundant features are difficult to identify accurately with a small number of training samples, but we can instead adjust feature magnitude with a soft mask based on estimated feature importance. We show that this method can generally improve few-shot transfer performance across various pretrained models and downstream datasets.
Across various sectors such as healthcare, criminal justice, national security, finance, and technology, large-scale machine learning (ML) and artificial intelligence (AI) systems are being deployed to make critical data-driven decisions. Many have asked if we can and should trust these ML systems to be making these decisions. Two critical components are prerequisites for trust in ML systems: interpretability, or the ability to understand why the ML system makes the decisions it does, and fairness, which ensures that ML systems do not exhibit bias against certain individuals or groups. Both interpretability and fairness are important and have separately received abundant attention in the ML literature, but so far, there have been very few methods developed to directly interpret models with regard to their fairness. In this paper, we focus on arguably the most popular type of ML interpretation: feature importance scores. Inspired by the use of decision trees in knowledge distillation, we propose to leverage trees as interpretable surrogates for complex black-box ML models. Specifically, we develop a novel fair feature importance score for trees that can be used to interpret how each feature contributes to fairness or bias in trees, tree-based ensembles, or tree-based surrogates of any complex ML system. Like the popular mean decrease in impurity for trees, our Fair Feature Importance Score is defined based on the mean decrease (or increase) in group bias. Through simulations as well as real examples on benchmark fairness datasets, we demonstrate that our Fair Feature Importance Score offers valid interpretations for both tree-based ensembles and tree-based surrogates of other ML systems.
The recent criticisms of the robustness of post hoc model approximation explanation methods (like LIME and SHAP) have led to the rise of model-precise abductive explanations. For each data point, abductive explanations provide a minimal subset of features that are sufficient to generate the outcome. While theoretically sound and rigorous, abductive explanations suffer from a major issue -- there can be several valid abductive explanations for the same data point. In such cases, providing a single abductive explanation can be insufficient; on the other hand, providing all valid abductive explanations can be incomprehensible due to their size. In this work, we solve this issue by aggregating the many possible abductive explanations into feature importance scores. We propose three aggregation methods: two based on power indices from cooperative game theory and a third based on a well-known measure of causal strength. We characterize these three methods axiomatically, showing that each of them uniquely satisfies a set of desirable properties. We also evaluate them on multiple datasets and show that these explanations are robust to the attacks that fool SHAP and LIME.
Heavy metal exposure is a common risk factor for hypertension. To develop an interpretable predictive machine learning (ML) model for hypertension based on levels of heavy metal exposure, data from the NHANES (2003–2016) were employed. Random forest (RF), support vector machine (SVM), decision tree (DT), multilayer perceptron (MLP), ridge regression (RR), AdaBoost (AB), gradient boosting decision tree (GBDT), voting classifier (VC), and K-nearest neighbour (KNN) algorithms were utilized to generate an optimal predictive model for hypertension. Three interpretable methods, the permutation feature importance analysis, partial dependence plot (PDP), and Shapley additive explanations (SHAP) methods, were integrated into a pipeline and embedded in ML for model interpretation. A total of 9005 eligible individuals were randomly allocated into two distinct sets for predictive model training and validation. The results showed that among the predictive models, the RF model demonstrated the highest performance, achieving an accuracy rate of 77.40% in the validation set. The AUC and F1 score for the model were 0.84 and 0.76, respectively. Blood Pb, urinary Cd, urinary Tl, and urinary Co levels were identified as the main influencers of hypertension, and their contribution weights were 0.0504 ± 0.0482, 0.0389 ± 0.0256, 0.0307 ± 0.0179, and 0.0296 ± 0.0162, respectively. Blood Pb (0.55–2.93 μg/dL) and urinary Cd (0.06–0.15 μg/L) levels exhibited the most pronounced upwards trend with the risk of hypertension within a specific value range, while urinary Tl (0.06–0.26 μg/L) and urinary Co (0.02–0.32 μg/L) levels demonstrated a declining trend with hypertension. The findings on the synergistic effects indicated that Pb and Cd were the primary determinants of hypertension. Our findings underscore the predictive value of heavy metals for hy pertension. By utilizing interpretable methods, we discerned that Pb, Cd, Tl, and Co emerged as noteworthy contributors within the predictive model.
Understanding the key factors shaping environmental hazard exposures and their associated environmental injustice issues is vital for formulating equitable policy measures. Traditional perspectives on environmental injustice have primarily focused on the socioeconomic dimensions, often overlooking the influence of heterogeneous urban characteristics. This limited view may obstruct a comprehensive understanding of the complex nature of environmental justice and its relationship with urban design features. To address this gap, this study creates an interpretable machine learning model to examine the effects of various urban features and their non-linear interactions to the exposure disparities of three primary hazards: air pollution, urban heat, and flooding. The analysis trains and tests models with data from six metropolitan counties in the United States using Random Forest and XGBoost. The performance is used to measure the extent to which variations of urban features shape disparities in environmental hazard levels. In addition, the analysis of feature importance reveals features related to social-demographic characteristics as the most prominent urban features that shape hazard extent. Features related to infrastructure distribution and land cover are relatively important for urban heat and air pollution exposure respectively. Moreover, we evaluate the models' transferability across different regions and hazards. The results highlight limited transferability, underscoring the intricate differences among hazards and regions and the way in which urban features shape hazard exposures. The insights gleaned from this study offer fresh perspectives on the relationship among urban features and their interplay with environmental hazard exposure disparities, informing the development of more integrated urban design policies to enhance social equity and environmental injustice issues.
Time series forecasting presents a significant challenge, particularly when its accuracy relies on external data sources rather than solely on historical values. This issue is prevalent in the financial sector, where the future behavior of time series is often intricately linked to information derived from various textual reports and a multitude of economic indicators. In practice, the key challenge lies in constructing a reliable time series forecasting model capable of harnessing data from diverse sources and extracting valuable insights to predict the target time series accurately. In this work, we tackle this challenging problem and introduce a novel multimodal transformer-based model named the \textit{Modality-aware Transformer}. Our model excels in exploring the power of both categorical text and numerical timeseries to forecast the target time series effectively while providing insights through its neural attention mechanism. To achieve this, we develop feature-level attention layers that encourage the model to focus on the most relevant features within each data modality. By incorporating the proposed feature-level attention, we develop a novel Intra-modal multi-head attention (MHA), Inter-modal MHA and Target-modal MHA in a way that both feature and temporal attentions are incorporated in MHAs. This enables the MHAs to generate temporal attentions with consideration of modality and feature importance which leads to more informative embeddings. The proposed modality-aware structure enables the model to effectively exploit information within each modality as well as foster cross-modal understanding. Our extensive experiments on financial datasets demonstrate that Modality-aware Transformer outperforms existing methods, offering a novel and practical solution to the complex challenges of multi-modal financial time series forecasting.
Anomaly detection is a critical task that involves the identification of data points that deviate from a predefined pattern, useful for fraud detection and related activities. Various techniques are employed for anomaly detection, but recent research indicates that deep learning methods, with their ability to discern intricate data patterns, are well-suited for this task. This study explores the use of Generative Adversarial Networks (GANs) for anomaly detection in power generation plants. The dataset used in this investigation comprises fuel consumption records obtained from power generation plants operated by a telecommunications company. The data was initially collected in response to observed irregularities in the fuel consumption patterns of the generating sets situated at the company's base stations. The dataset was divided into anomalous and normal data points based on specific variables, with 64.88% classified as normal and 35.12% as anomalous. An analysis of feature importance, employing the random forest classifier, revealed that Running Time Per Day exhibited the highest relative importance. A GANs model was trained and fine-tuned both with and without data augmentation, with the goal of increasing the dataset size to enhance performance. The generator model consisted of five dense layers using the tanh activation function, while the discriminator comprised six dense layers, each integrated with a dropout layer to prevent overfitting. Following data augmentation, the model achieved an accuracy rate of 98.99%, compared to 66.45% before augmentation. This demonstrates that the model nearly perfectly classified data points into normal and anomalous categories, with the augmented data significantly enhancing the GANs' performance in anomaly detection. Consequently, this study recommends the use of GANs, particularly when using large datasets, for effective anomaly detection.
The aim of this study is to analyze the effect of serum metabolites on diabetic nephropathy (DN) and predict the prevalence of DN through a machine learning approach. The dataset consists of 548 patients from April 2018 to April 2019 in Second Affiliated Hospital of Dalian Medical University (SAHDMU). We select the optimal 38 features through a Least absolute shrinkage and selection operator (LASSO) regression model and a 10-fold cross-validation. We compare four machine learning algorithms, including eXtreme Gradient Boosting (XGB), random forest, decision tree and logistic regression, by AUC-ROC curves, decision curves, calibration curves. We quantify feature importance and interaction effects in the optimal predictive model by Shapley Additive exPlanations (SHAP) method. The XGB model has the best performance to screen for DN with the highest AUC value of 0.966. The XGB model also gains more clinical net benefits than others and the fitting degree is better. In addition, there are significant interactions between serum metabolites and duration of diabetes. We develop a predictive model by XGB algorithm to screen for DN. C2, C5DC, Tyr, Ser, Met, C24, C4DC, and Cys have great contribution in the model, and can possibly be biomarkers for DN.