PyTorchKR Papers

최신 논문

총 848편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2104.07377

Facilitating Machine Learning Model Comparison and Explanation Through A Radial Visualisation

Building an effective Machine Learning (ML) model for a data set is a difficult task involving various steps. One of the most important steps is to compare generated substantial amounts of ML models to find the optimal one for the deployment. It is challenging to compare such models with dynamic number of features. Comparison is more than just finding differences of ML model performance, users are also interested in the relations between features and model performance such as feature importance for ML explanations. This paper proposes RadialNet Chart, a novel visualisation approach to compare ML models trained with a different number of features of a given data set while revealing implicit dependent relations. In RadialNet Chart, ML models and features are represented by lines and arcs respectively. These lines are generated effectively using a recursive function. The dependence of ML models with dynamic number of features is encoded into the structure of visualisation, where ML models and their dependent features are directly revealed from related line connections. ML model performance information is encoded with colour and line width in RadialNet Chart. Together with the structure of visualisation, feature importance can be directly discerned in RadialNet Chart for ML explanations.

arXiv:2104.07225공식 코드 · Papers with Code

Text Guide: Improving the quality of long text classification by a text selection method based on feature importance

The performance of text classification methods has improved greatly over the last decade for text instances of less than 512 tokens. This limit has been adopted by most state-of-the-research transformer models due to the high computational cost of analyzing longer text instances. To mitigate this problem and to improve classification for longer texts, researchers have sought to resolve the underlying causes of the computational cost and have proposed optimizations for the attention mechanism, which is the key element of every transformer model. In our study, we are not pursuing the ultimate goal of long text classification, i.e., the ability to analyze entire text instances at one time while preserving high performance at a reasonable computational cost. Instead, we propose a text truncation method called Text Guide, in which the original text length is reduced to a predefined limit in a manner that improves performance over naive and semi-naive approaches while preserving low computational costs. Text Guide benefits from the concept of feature importance, a notion from the explainable artificial intelligence domain. We demonstrate that Text Guide can be used to improve the performance of recent language models specifically designed for long text classification, such as Longformer. Moreover, we discovered that parameter optimization is the key to Text Guide performance and must be conducted before the method is deployed. Future experiments may reveal additional benefits provided by this new method.

arXiv:2104.06057공식 코드 · Papers with Code

LioNets: A Neural-Specific Local Interpretation Technique Exploiting Penultimate Layer Information

Artificial Intelligence (AI) has a tremendous impact on the unexpected growth of technology in almost every aspect. AI-powered systems are monitoring and deciding about sensitive economic and societal issues. The future is towards automation, and it must not be prevented. However, this is a conflicting viewpoint for a lot of people, due to the fear of uncontrollable AI systems. This concern could be reasonable if it was originating from considerations associated with social issues, like gender-biased, or obscure decision-making systems. Explainable AI (XAI) is recently treated as a huge step towards reliable systems, enhancing the trust of people to AI. Interpretable machine learning (IML), a subfield of XAI, is also an urgent topic of research. This paper presents a small but significant contribution to the IML community, focusing on a local-based, neural-specific interpretation process applied to textual and time-series data. The proposed methodology introduces new approaches to the presentation of feature importance based interpretations, as well as the production of counterfactual words on textual datasets. Eventually, an improved evaluation metric is introduced for the assessment of interpretation techniques, which supports an extensive set of qualitative and quantitative experiments.

arXiv:2104.04295공식 코드 · Papers with Code

Transforming Feature Space to Interpret Machine Learning Models

Model-agnostic tools for interpreting machine-learning models struggle to summarize the joint effects of strongly dependent features in high-dimensional feature spaces, which play an important role in pattern recognition, for example in remote sensing of landcover. This contribution proposes a novel approach that interprets machine-learning models through the lens of feature space transformations. It can be used to enhance unconditional as well as conditional post-hoc diagnostic tools including partial dependence plots, accumulated local effects plots, or permutation feature importance assessments. While the approach can also be applied to nonlinear transformations, we focus on linear ones, including principal component analysis (PCA) and a partial orthogonalization technique. Structured PCA and diagnostics along paths offer opportunities for representing domain knowledge. The new approach is implemented in the R package `wiml`, which can be combined with existing explainable machine-learning packages. A case study on remote-sensing landcover classification with 46 features is used to demonstrate the potential of the proposed approach for model interpretation by domain experts.

arXiv:2104.04375

Model LineUpper: Supporting Interactive Model Comparison at Multiple Levels for AutoML

Automated Machine Learning (AutoML) is a rapidly growing set of technologies that automate the model development pipeline by searching model space and generating candidate models. A critical, final step of AutoML is human selection of a final model from dozens of candidates. In current AutoML systems, selection is supported only by performance metrics. Prior work has shown that in practice, people evaluate ML models based on additional criteria, such as the way a model makes predictions. Comparison may happen at multiple levels, from types of errors, to feature importance, to how the model makes predictions of specific instances. We developed \tool{} to support interactive model comparison for AutoML by integrating multiple Explainable AI (XAI) and visualization techniques. We conducted a user study in which we both evaluated the system and used it as a technology probe to understand how users perform model comparison in an AutoML system. We discuss design implications for utilizing XAI techniques for model comparison and supporting the unique needs of data scientists in comparing AutoML models.

arXiv:2104.03088

Hollow-tree Super: a directional and scalable approach for feature importance in boosted tree models

Current limitations in boosted tree modelling prevent the effective scaling to datasets with a large feature number, particularly when investigating the magnitude and directionality of various features on classification. We present a novel methodology, Hollow-tree Super (HOTS), to resolve and visualize feature importance in boosted tree models involving a large number of features. Further, HOTS allows for investigation of the directionality and magnitude various features have on classification. Using the Iris dataset, we first compare HOTS to Gini Importance, Partial Dependence Plots, and Permutation Importance, and demonstrate how HOTS resolves the weaknesses present in these methods. We then show how HOTS can be utilized in high dimensional neuroscientific data, by taking 60 Schizophrenic subjects and applying the method to determine which brain regions were most important for classification of schizophrenia as determined by the PANSS. HOTS effectively replicated and supported the findings of Gini importance, Partial Dependence Plots and Permutation importance within the Iris dataset. When applied to the schizophrenic brain dataset, HOTS was able to resolve the top 10 most important features for classification, as well as their directionality for classification and magnitude compared to other features. Cross-validation supported that these same 10 features were consistently used in the decision-making process across multiple trees, and these features were localised primarily to the occipital and parietal cortices, commonly disturbed brain regions in those with Schizophrenia. It is imperative that a methodology is developed that is able to handle the demands of working with large datasets that contain a large number of features. HOTS represents a unique way to investigate both the directionality and magnitude of feature importance when working at scale with boosted-tree modelling.

arXiv:2104.02883

Online Feature Screening for Data Streams with Concept Drift

Screening feature selection methods are often used as a preprocessing step for reducing the number of variables before training step. Traditional screening methods only focus on dealing with complete high dimensional datasets. Modern datasets not only have higher dimension and larger sample size, but also have properties such as streaming input, sparsity and concept drift. Therefore a considerable number of online feature selection methods were introduced to handle these kind of problems in recent years. Online screening methods are one of the categories of online feature selection methods. The methods that we proposed in this research are capable of handling all three situations mentioned above. Our research study focuses on classification datasets. Our experiments show proposed methods can generate the same feature importance as their offline version with faster speed and less storage consumption. Furthermore, the results show that online screening methods with integrated model adaptation have a higher true feature detection rate than without model adaptation on data streams with the concept drift property. Among the two large real datasets that potentially have the concept drift property, online screening methods with model adaptation show advantages in either saving computing time and space, reducing model complexity, or improving prediction accuracy.

arXiv:2104.00411공식 코드 · Papers with Code

Explaining COVID-19 and Thoracic Pathology Model Predictions by Identifying Informative Input Features

Neural networks have demonstrated remarkable performance in classification and regression tasks on chest X-rays. In order to establish trust in the clinical routine, the networks' prediction mechanism needs to be interpretable. One principal approach to interpretation is feature attribution. Feature attribution methods identify the importance of input features for the output prediction. Building on Information Bottleneck Attribution (IBA) method, for each prediction we identify the chest X-ray regions that have high mutual information with the network's output. Original IBA identifies input regions that have sufficient predictive information. We propose Inverse IBA to identify all informative regions. Thus all predictive cues for pathologies are highlighted on the X-rays, a desirable property for chest X-ray diagnosis. Moreover, we propose Regression IBA for explaining regression models. Using Regression IBA we observe that a model trained on cumulative severity score labels implicitly learns the severity of different X-ray regions. Finally, we propose Multi-layer IBA to generate higher resolution and more detailed attribution/saliency maps. We evaluate our methods using both human-centric (ground-truth-based) interpretability metrics, and human-independent feature importance metrics on NIH Chest X-ray8 and BrixIA datasets. The Code is publicly available.

arXiv:2103.11678

Feature Selection for Imbalanced Data with Deep Sparse Autoencoders Ensemble

Class imbalance is a common issue in many domain applications of learning algorithms. Oftentimes, in the same domains it is much more relevant to correctly classify and profile minority class observations. This need can be addressed by Feature Selection (FS), that offers several further advantages, s.a. decreasing computational costs, aiding inference and interpretability. However, traditional FS techniques may become sub-optimal in the presence of strongly imbalanced data. To achieve FS advantages in this setting, we propose a filtering FS algorithm ranking feature importance on the basis of the Reconstruction Error of a Deep Sparse AutoEncoders Ensemble (DSAEE). We use each DSAE trained only on majority class to reconstruct both classes. From the analysis of the aggregated Reconstruction Error, we determine the features where the minority class presents a different distribution of values w.r.t. the overrepresented one, thus identifying the most relevant features to discriminate between the two. We empirically demonstrate the efficacy of our algorithm in several experiments on high-dimensional datasets of varying sample size, showcasing its capability to select relevant and generalizable features to profile and classify minority class, outperforming other benchmark FS methods. We also briefly present a real application in radiogenomics, where the methodology was applied successfully.

arXiv:2103.07765

Image Classifiers for Network Intrusions

This research recasts the network attack dataset from UNSW-NB15 as an intrusion detection problem in image space. Using one-hot-encodings, the resulting grayscale thumbnails provide a quarter-million examples for deep learning algorithms. Applying the MobileNetV2's convolutional neural network architecture, the work demonstrates a 97% accuracy in distinguishing normal and attack traffic. Further class refinements to 9 individual attack families (exploits, worms, shellcodes) show an overall 56% accuracy. Using feature importance rank, a random forest solution on subsets show the most important source-destination factors and the least important ones as mainly obscure protocols. The dataset is available on Kaggle.

arXiv:2103.06397

Interpretable Data-driven Methods for Subgrid-scale Closure in LES for Transcritical LOX/GCH4 Combustion

Many practical combustion systems such as those in rockets, gas turbines, and internal combustion engines operate under high pressures that surpass the thermodynamic critical limit of fuel-oxidizer mixtures. These conditions require the consideration of complex fluid behaviors that pose challenges for numerical simulations, casting doubts on the validity of existing subgrid-scale (SGS) models in large-eddy simulations of these systems. While data-driven methods have shown high accuracy as closure models in simulations of turbulent flames, these models are often criticized for lack of physical interpretability, wherein they provide answers but no insight into their underlying rationale. The objective of this study is to assess SGS stress models from conventional physics-driven approaches and an interpretable machine learning algorithm, i.e., the random forest regressor, in a turbulent transcritical non-premixed flame. To this end, direct numerical simulations (DNS) of transcritical liquid-oxygen/gaseous-methane (LOX/GCH4) inert and reacting flows are performed. Using this data, a priori analysis is performed on the Favre-filtered DNS data to examine the accuracy of physics-based and random forest SGS-models under these conditions. SGS stresses calculated with the gradient model show good agreement with the exact terms extracted from filtered DNS. The accuracy of the random-forest regressor decreased when physics-based constraints are applied to the feature set. Results demonstrate that random forests can perform as effectively as algebraic models when modeling subgrid stresses, only when trained on a sufficiently representative database. The employment of random forest feature importance score is shown to provide insight into discovering subgrid-scale stresses through sparse regression.

arXiv:2103.04951

A Comparative Approach to Explainable Artificial Intelligence Methods in Application to High-Dimensional Electronic Health Records: Examining the Usability of XAI

Explainable Artificial Intelligence (XAI) is a rising field in AI. It aims to produce a demonstrative factor of trust, which for human subjects is achieved through communicative means, which Machine Learning (ML) algorithms cannot solely produce, illustrating the necessity of an extra layer producing support to the model output. When approaching the medical field, we can see challenges arise when dealing with the involvement of human-subjects, the ideology behind trusting a machine to tend towards the livelihood of a human poses an ethical conundrum - leaving trust as the basis of the human-expert in acceptance to the machines decision. The aim of this paper is to apply XAI methods to demonstrate the usability of explainable architectures as a tertiary layer for the medical domain supporting ML predictions and human-expert opinion, XAI methods produce visualization of the feature contribution towards a given models output on both a local and global level. The work in this paper uses XAI to determine feature importance towards high-dimensional data-driven questions to inform domain-experts of identifiable trends with a comparison of model-agnostic methods in application to ML algorithms. The performance metrics for a glass-box method is also provided as a comparison against black-box capability for tabular data. Future work will aim to produce a user-study using metrics to evaluate human-expert usability and opinion of the given models.

arXiv:2103.01300

Understanding & Predicting User Lifetime with Machine Learning in an Anonymous Location-Based Social Network

In this work, we predict the user lifetime within the anonymous and location-based social network Jodel in the Kingdom of Saudi Arabia. Jodel's location-based nature yields to the establishment of disjoint communities country-wide and enables for the first time the study of user lifetime in the case of a large set of disjoint communities. A user's lifetime is an important measurement for evaluating and steering customer bases as it can be leveraged to predict churn and possibly apply suitable methods to circumvent potential user losses. We train and test off the shelf machine learning techniques with 5-fold crossvalidation to predict user lifetime as a regression and classification problem; identifying the Random Forest to provide very strong results. Discussing model complexity and quality trade-offs, we also dive deep into a time-dependent feature subset analysis, which does not work very well; Easing up the classification problem into a binary decision (lifetime longer than timespan $x$) enables a practical lifetime predictor with very good performance. We identify implicit similarities across community models according to strong correlations in feature importance. A single countrywide model generalizes the problem and works equally well for any tested community; the overall model internally works similar to others also indicated by its feature importances.

arXiv:2102.11934공식 코드 · Papers with Code

Feature Importance Explanations for Temporal Black-Box Models

Models in the supervised learning framework may capture rich and complex representations over the features that are hard for humans to interpret. Existing methods to explain such models are often specific to architectures and data where the features do not have a time-varying component. In this work, we propose TIME, a method to explain models that are inherently temporal in nature. Our approach (i) uses a model-agnostic permutation-based approach to analyze global feature importance, (ii) identifies the importance of salient features with respect to their temporal ordering as well as localized windows of influence, and (iii) uses hypothesis testing to provide statistical rigor.

arXiv:2102.11848

An Explainable Artificial Intelligence Approach for Unsupervised Fault Detection and Diagnosis in Rotating Machinery

The monitoring of rotating machinery is an essential task in today's production processes. Currently, several machine learning and deep learning-based modules have achieved excellent results in fault detection and diagnosis. Nevertheless, to further increase user adoption and diffusion of such technologies, users and human experts must be provided with explanations and insights by the modules. Another issue is related, in most cases, with the unavailability of labeled historical data that makes the use of supervised models unfeasible. Therefore, a new approach for fault detection and diagnosis in rotating machinery is here proposed. The methodology consists of three parts: feature extraction, fault detection and fault diagnosis. In the first part, the vibration features in the time and frequency domains are extracted. Secondly, in the fault detection, the presence of fault is verified in an unsupervised manner based on anomaly detection algorithms. The modularity of the methodology allows different algorithms to be implemented. Finally, in fault diagnosis, Shapley Additive Explanations (SHAP), a technique to interpret black-box models, is used. Through the feature importance ranking obtained by the model explainability, the fault diagnosis is performed. Two tools for diagnosis are proposed, namely: unsupervised classification and root cause analysis. The effectiveness of the proposed approach is shown on three datasets containing different mechanical faults in rotating machinery. The study also presents a comparison between models used in machine learning explainability: SHAP and Local Depth-based Feature Importance for the Isolation Forest (Local- DIFFI). Lastly, an analysis of several state-of-art anomaly detection algorithms in rotating machinery is included.

arXiv:2102.10284공식 코드 · Papers with Code

Artificial Intelligence Enhanced Rapid and Efficient Diagnosis of Mycoplasma Pneumoniae Pneumonia in Children Patients

Artificial intelligence methods have been increasingly turning into a potentially powerful tool in the diagnosis and management of diseases. In this study, we utilized logistic regression (LR), decision tree (DT), gradient boosted decision tree (GBDT), support vector machine (SVM), and multilayer perceptron (MLP) as machine learning models to rapidly diagnose the mycoplasma pneumoniae pneumonia (MPP) in children patients. The classification task was carried out after applying the preprocessing procedure to the MPP dataset. The most efficient results are obtained by GBDT. It provides the best performance with an accuracy of 93.7%. In contrast to standard raw feature weighting, the feature importance takes the underlying correlation structure of the features into account. The most crucial feature of GBDT is the "pulmonary infiltrates range" with a score of 0.5925, followed by "cough" (0.0953) and "pleural effusion" (0.0492). We publicly share our full implementation with the dataset and trained models at https://github.com/zhenguonie/2021_AI4MPP.

arXiv:2102.08540

Intuitively Assessing ML Model Reliability through Example-Based Explanations and Editing Model Inputs

Interpretability methods aim to help users build trust in and understand the capabilities of machine learning models. However, existing approaches often rely on abstract, complex visualizations that poorly map to the task at hand or require non-trivial ML expertise to interpret. Here, we present two visual analytics modules that facilitate an intuitive assessment of model reliability. To help users better characterize and reason about a model's uncertainty, we visualize raw and aggregate information about a given input's nearest neighbors. Using an interactive editor, users can manipulate this input in semantically-meaningful ways, determine the effect on the output, and compare against their prior expectations. We evaluate our interface using an electrocardiogram beat classification case study. Compared to a baseline feature importance interface, we find that 14 physicians are better able to align the model's uncertainty with domain-relevant factors and build intuition about its capabilities and limitations.

arXiv:2102.06761

MIMIC-IF: Interpretability and Fairness Evaluation of Deep Learning Models on MIMIC-IV Dataset

The recent release of large-scale healthcare datasets has greatly propelled the research of data-driven deep learning models for healthcare applications. However, due to the nature of such deep black-boxed models, concerns about interpretability, fairness, and biases in healthcare scenarios where human lives are at stake call for a careful and thorough examinations of both datasets and models. In this work, we focus on MIMIC-IV (Medical Information Mart for Intensive Care, version IV), the largest publicly available healthcare dataset, and conduct comprehensive analyses of dataset representation bias as well as interpretability and prediction fairness of deep learning models for in-hospital mortality prediction. In terms of interpretabilty, we observe that (1) the best performing interpretability method successfully identifies critical features for mortality prediction on various prediction models; (2) demographic features are important for prediction. In terms of fairness, we observe that (1) there exists disparate treatment in prescribing mechanical ventilation among patient groups across ethnicity, gender and age; (2) all of the studied mortality predictors are generally fair while the IMV-LSTM (Interpretable Multi-Variable Long Short-Term Memory) model provides the most accurate and unbiased predictions across all protected groups. We further draw concrete connections between interpretability methods and fairness metrics by showing how feature importance from interpretability methods can be beneficial in quantifying potential disparities in mortality predictors.

arXiv:2102.06029

Feature Analyses and Modelling of Lithium-ion Batteries Manufacturing based on Random Forest Classification

Lithium-ion battery manufacturing is a highly complicated process with strongly coupled feature interdependencies, a feasible solution that can analyse feature variables within manufacturing chain and achieve reliable classification is thus urgently needed. This article proposes a random forest (RF)-based classification framework, through using the out of bag (OOB) predictions, Gini changes as well as predictive measure of association (PMOA), for effectively quantifying the importance and correlations of battery manufacturing features and their effects on the classification of electrode properties. Battery manufacturing data containing three intermediate product features from the mixing stage and one product parameter from the coating stage are analysed by the designed RF framework to investigate their effects on both the battery electrode active material mass load and porosity. Illustrative results demonstrate that the proposed RF framework not only achieves the reliable classification of electrode properties but also leads to the effective quantification of both manufacturing feature importance and correlations. This is the first time to design a systematic RF framework for simultaneously quantifying battery production feature importance and correlations by three various quantitative indicators including the unbiased feature importance (FI), gain improvement FI and PMOA, paving a promising solution to reduce model dimension and conduct efficient sensitivity analysis of battery manufacturing.

arXiv:2102.05228

Towards Better Explanations of Class Activation Mapping

Increasing demands for understanding the internal behavior of convolutional neural networks (CNNs) have led to remarkable improvements in explanation methods. Particularly, several class activation mapping (CAM) based methods, which generate visual explanation maps by a linear combination of activation maps from CNNs, have been proposed. However, the majority of the methods lack a clear theoretical basis on how they assign the coefficients of the linear combination. In this paper, we revisit the intrinsic linearity of CAM with respect to the activation maps; we construct an explanation model of CNN as a linear function of binary variables that denote the existence of the corresponding activation maps. With this approach, the explanation model can be determined by additive feature attribution methods in an analytic manner. We then demonstrate the adequacy of SHAP values, which is a unique solution for the explanation model with a set of desirable properties, as the coefficients of CAM. Since the exact SHAP values are unattainable, we introduce an efficient approximation method, LIFT-CAM, based on DeepLIFT. Our proposed LIFT-CAM can estimate the SHAP values of the activation maps with high speed and accuracy. Furthermore, it greatly outperforms other previous CAM-based methods in both qualitative and quantitative aspects.