PyTorchKR Papers

최신 논문

총 9,382편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2206.02034

A Control Theoretic Framework for Adaptive Gradient Optimizers in Machine Learning

Adaptive gradient methods have become popular in optimizing deep neural networks; recent examples include AdaGrad and Adam. Although Adam usually converges faster, variations of Adam, for instance, the AdaBelief algorithm, have been proposed to enhance Adam's poor generalization ability compared to the classical stochastic gradient method. This paper develops a generic framework for adaptive gradient methods that solve non-convex optimization problems. We first model the adaptive gradient methods in a state-space framework, which allows us to present simpler convergence proofs of adaptive optimizers such as AdaGrad, Adam, and AdaBelief. We then utilize the transfer function paradigm from classical control theory to propose a new variant of Adam, coined AdamSSM. We add an appropriate pole-zero pair in the transfer function from squared gradients to the second moment estimate. We prove the convergence of the proposed AdamSSM algorithm. Applications on benchmark machine learning tasks of image classification using CNN architectures and language modeling using LSTM architecture demonstrate that the AdamSSM algorithm improves the gap between generalization accuracy and faster convergence than the recent adaptive gradient methods.

arXiv:2206.01906

Distributed Machine Learning in D2D-Enabled Heterogeneous Networks: Architectures, Performance, and Open Challenges

The ever-growing concerns regarding data privacy have led to a paradigm shift in machine learning (ML) architectures from centralized to distributed approaches, giving rise to federated learning (FL) and split learning (SL) as the two predominant privacy-preserving ML mechanisms. However,implementing FL or SL in device-to-device (D2D)-enabled heterogeneous networks with diverse clients presents substantial challenges, including architecture scalability and prolonged training delays. To address these challenges, this article introduces two innovative hybrid distributed ML architectures, namely, hybrid split FL (HSFL) and hybrid federated SL (HFSL). Such architectures combine the strengths of both FL and SL in D2D-enabled heterogeneous wireless networks. We provide a comprehensive analysis of the performance and advantages of HSFL and HFSL, while also highlighting open challenges for future exploration. We support our proposals with preliminary simulations using three datasets in non-independent and non-identically distributed settings, demonstrating the feasibility of our architectures. Our simulations reveal notable reductions in communication/computation costs and training delays as compared to conventional FL and SL.

arXiv:2206.02671

Canonical Cortical Graph Neural Networks and its Application for Speech Enhancement in Audio-Visual Hearing Aids

Despite the recent success of machine learning algorithms, most models face drawbacks when considering more complex tasks requiring interaction between different sources, such as multimodal input data and logical time sequences. On the other hand, the biological brain is highly sharpened in this sense, empowered to automatically manage and integrate such streams of information. In this context, this work draws inspiration from recent discoveries in brain cortical circuits to propose a more biologically plausible self-supervised machine learning approach. This combines multimodal information using intra-layer modulations together with Canonical Correlation Analysis, and a memory mechanism to keep track of temporal data, the overall approach termed Canonical Cortical Graph Neural networks. This is shown to outperform recent state-of-the-art models in terms of clean audio reconstruction and energy efficiency for a benchmark audio-visual speech dataset. The enhanced performance is demonstrated through a reduced and smother neuron firing rate distribution. suggesting that the proposed model is amenable for speech enhancement in future audio-visual hearing aid devices.

arXiv:2206.01728

A review of machine learning approaches, challenges and prospects for computational tumor pathology

Computational pathology is part of precision oncology medicine. The integration of high-throughput data including genomics, transcriptomics, proteomics, metabolomics, pathomics, and radiomics into clinical practice improves cancer treatment plans, treatment cycles, and cure rates, and helps doctors open up innovative approaches to patient prognosis. In the past decade, rapid advances in artificial intelligence, chip design and manufacturing, and mobile computing have facilitated research in computational pathology and have the potential to provide better-integrated solutions for whole-slide images, multi-omics data, and clinical informatics. However, tumor computational pathology now brings some challenges to the application of tumour screening, diagnosis and prognosis in terms of data integration, hardware processing, network sharing bandwidth and machine learning technology. This review investigates image preprocessing methods in computational pathology from a pathological and technical perspective, machine learning-based methods, and applications of computational pathology in breast, colon, prostate, lung, and various tumour disease scenarios. Finally, the challenges and prospects of machine learning in computational pathology applications are discussed.

arXiv:2206.02511

Transfer Learning based Search Space Design for Hyperparameter Tuning

The tuning of hyperparameters becomes increasingly important as machine learning (ML) models have been extensively applied in data mining applications. Among various approaches, Bayesian optimization (BO) is a successful methodology to tune hyper-parameters automatically. While traditional methods optimize each tuning task in isolation, there has been recent interest in speeding up BO by transferring knowledge across previous tasks. In this work, we introduce an automatic method to design the BO search space with the aid of tuning history from past tasks. This simple yet effective approach can be used to endow many existing BO methods with transfer learning capabilities. In addition, it enjoys the three advantages: universality, generality, and safeness. The extensive experiments show that our approach considerably boosts BO by designing a promising and compact search space instead of using the entire space, and outperforms the state-of-the-arts on a wide range of benchmarks, including machine learning and deep learning tuning tasks, and neural architecture search.

arXiv:2206.02409

Is More Data All You Need? A Causal Exploration

Curating a large scale medical imaging dataset for machine learning applications is both time consuming and expensive. Balancing the workload between model development, data collection and annotations is difficult for machine learning practitioners, especially under time constraints. Causal analysis is often used in medicine and economics to gain insights about the effects of actions and policies. In this paper we explore the effect of dataset interventions on the output of image classification models. Through a causal approach we investigate the effects of the quantity and type of data we need to incorporate in a dataset to achieve better performance for specific subtasks. The main goal of this paper is to highlight the potential of causal analysis as a tool for resource optimization for developing medical imaging ML applications. We explore this concept with a synthetic dataset and an exemplary use-case for Diabetic Retinopathy image analysis.

arXiv:2206.01749

Uncertainty Estimation in Machine Learning

Most machine learning techniques are based upon statistical learning theory, often simplified for the sake of computing speed. This paper is focused on the uncertainty aspect of mathematical modeling in machine learning. Regression analysis is chosen to further investigate the evaluation aspect of uncertainty in model coefficients and, more importantly, in the output feature value predictions. A survey demonstrates major stages in the conventional least squares approach to the creation of the regression model, along with its uncertainty estimation. On the other hand, it is shown that in machine learning the model complexity and severe nonlinearity become serious obstacles to uncertainty evaluation. Furthermore, the process of machine model training demands high computing power, not available at the level of personal computers. This is why so-called pre-trained models are widely used in such areas of machine learning as natural language processing. The latest example of a pre-trained model is the Generative Pre-trained Transformer 3 with hundreds of billions of parameters and a half-terabyte training dataset. Similarly, mathematical models built from real data are growing in complexity which is accompanied by the growing amount of training data. However, when machine models and their predictions are used in decision-making, one needs to estimate uncertainty and evaluate accompanying risks. This problem could be resolved with non-parametric techniques at the expense of greater demand for computing power, which can be offered by modern supercomputers available, including those utilizing graphical and tensor processing units along with the conventional central processors.

arXiv:2206.01538

Accelerating hydrodynamic simulations of urban drainage systems with physics-guided machine learning

We propose and demonstrate a new approach for fast and accurate surrogate modelling of urban drainage system hydraulics based on physics-guided machine learning. The surrogates are trained against a limited set of simulation results from a hydrodynamic (HiFi) model. Our approach reduces simulation times by one to two orders of magnitude compared to a HiFi model. It is thus slower than e.g. conceptual hydrological models, but it enables simulations of water levels, flows and surcharges in all nodes and links of a drainage network and thus largely preserves the level of detail provided by HiFi models. Comparing time series simulated by the surrogate and the HiFi model, R2 values in the order of 0.9 are achieved. Surrogate training times are currently in the order of one hour. However, they can likely be reduced through the application of transfer learning and graph neural networks. Our surrogate approach will be useful for interactive workshops in initial design phases of urban drainage systems, as well as for real time applications. In addition, our model formulation is generic and future research should investigate its application for simulating other water systems.

arXiv:2206.01562

Prescriptive maintenance with causal machine learning

Machine maintenance is a challenging operational problem, where the goal is to plan sufficient preventive maintenance to avoid machine failures and overhauls. Maintenance is often imperfect in reality and does not make the asset as good as new. Although a variety of imperfect maintenance policies have been proposed in the literature, these rely on strong assumptions regarding the effect of maintenance on the machine's condition, assuming the effect is (1) deterministic or governed by a known probability distribution, and (2) machine-independent. This work proposes to relax both assumptions by learning the effect of maintenance conditional on a machine's characteristics from observational data on similar machines using existing methodologies for causal inference. By predicting the maintenance effect, we can estimate the number of overhauls and failures for different levels of maintenance and, consequently, optimize the preventive maintenance frequency to minimize the total estimated cost. We validate our proposed approach using real-life data on more than 4,000 maintenance contracts from an industrial partner. Empirical results show that our novel, causal approach accurately predicts the maintenance effect and results in individualized maintenance schedules that are more accurate and cost-effective than supervised or non-individualized approaches.

arXiv:2206.01523

A High-Performance Customer Churn Prediction System based on Self-Attention

Customer churn prediction is a challenging domain of research that contributes to customer retention strategy. The predictive performance of existing machine learning models, which are often adopted by churn communities, appear to be at a bottleneck, partly due to models' poor feature extraction capability. Therefore, a novel algorithm, a hybrid neural network with self-attention enhancement (HNNSAE), is proposed in this paper to improve the efficiency of feature screening and feature extraction, consequently improving the model's predictive performance. This model consists of three main blocks. The first block is the entity embedding layer, which is employed to process the categorical variables transformed into 0-1 code. The second block is the feature extractor, which extracts the significant features through the multi-head self-attention mechanism. In addition, to improve the feature extraction effect, we stack the residual connection neural network on multi-head self-attention modules. The third block is a classifier, which is a three-layer multilayer perceptron. This work conducts experiments on publicly available dataset related to commercial bank customers. The result demonstrates that HNNSAE significantly outperforms the other Individual Machine Learning (IML), Ensemble Machine Learning (EML), and Deep Learning (DL) methods tested in this paper. Furthermore, we compare the performance of the feature extractor proposed in this paper with that of other three feature extractors and find that the method proposed in this paper significantly outperforms other methods. In addition, four hypotheses about model prediction performance and overfitting risk are tested on the publicly available dataset.

arXiv:2206.01314

A New Security Boundary of Component Differentially Challenged XOR PUFs Against Machine Learning Modeling Attacks

Physical Unclonable Functions (PUFs) are promising security primitives for resource-constrained network nodes. The XOR Arbiter PUF (XOR PUF or XPUF) is an intensively studied PUF invented to improve the security of the Arbiter PUF, probably the most lightweight delay-based PUF. Recently, highly powerful machine learning attack methods were discovered and were able to easily break large-sized XPUFs, which were highly secure against earlier machine learning attack methods. Component-differentially-challenged XPUFs (CDC-XPUFs) are XPUFs with different component PUFs receiving different challenges. Studies showed they were much more secure against machine learning attacks than the conventional XPUFs, whose component PUFs receive the same challenge. But these studies were all based on earlier machine learning attack methods, and hence it is not clear if CDC-XPUFs can remain secure under the recently discovered powerful attack methods. In this paper, the two current most powerful two machine learning methods for attacking XPUFs are adapted by fine-tuning the parameters of the two methods for CDC-XPUFs. Attack experiments using both simulated PUF data and silicon data generated from PUFs implemented on field-programmable gate array (FPGA) were carried out, and the experimental results showed that some previously secure CDC-XPUFs of certain circuit parameter values are no longer secure under the adapted new attack methods, while many more CDC-XPUFs of other circuit parameter values remain secure. Thus, our experimental attack study has re-defined the boundary between the secure region and the insecure region of the PUF circuit parameter space, providing PUF manufacturers and IoT security application developers with valuable information in choosing PUFs with secure parameter values.

arXiv:2206.01640

PROMISSING: Pruning Missing Values in Neural Networks

While data are the primary fuel for machine learning models, they often suffer from missing values, especially when collected in real-world scenarios. However, many off-the-shelf machine learning models, including artificial neural network models, are unable to handle these missing values directly. Therefore, extra data preprocessing and curation steps, such as data imputation, are inevitable before learning and prediction processes. In this study, we propose a simple and intuitive yet effective method for pruning missing values (PROMISSING) during learning and inference steps in neural networks. In this method, there is no need to remove or impute the missing values; instead, the missing values are treated as a new source of information (representing what we do not know). Our experiments on simulated data, several classification and regression benchmarks, and a multi-modal clinical dataset show that PROMISSING results in similar prediction performance compared to various imputation techniques. In addition, our experiments show models trained using PROMISSING techniques are becoming less decisive in their predictions when facing incomplete samples with many unknowns. This finding hopefully advances machine learning models from being pure predicting machines to more realistic thinkers that can also say "I do not know" when facing incomplete sources of information.

arXiv:2206.00807

Applied Federated Learning: Architectural Design for Robust and Efficient Learning in Privacy Aware Settings

The classical machine learning paradigm requires the aggregation of user data in a central location where machine learning practitioners can preprocess data, calculate features, tune models and evaluate performance. The advantage of this approach includes leveraging high performance hardware (such as GPUs) and the ability of machine learning practitioners to do in depth data analysis to improve model performance. However, these advantages may come at a cost to data privacy. User data is collected, aggregated, and stored on centralized servers for model development. Centralization of data poses risks, including a heightened risk of internal and external security incidents as well as accidental data misuse. Federated learning with differential privacy is designed to avoid the server-side centralization pitfall by bringing the ML learning step to users' devices. Learning is done in a federated manner where each mobile device runs a training loop on a local copy of a model. Updates from on-device models are sent to the server via encrypted communication and through differential privacy to improve the global model. In this paradigm, users' personal data remains on their devices. Surprisingly, model training in this manner comes at a fairly minimal degradation in model performance. However, federated learning comes with many other challenges due to its distributed nature, heterogeneous compute environments and lack of data visibility. This paper explores those challenges and outlines an architectural design solution we are exploring and testing to productionize federated learning at Meta scale.

arXiv:2206.00727

(Machine) Learning What Policies Value

When a policy prioritizes one person over another, is it because they benefit more, or because they are preferred? This paper develops a method to uncover the values consistent with observed allocation decisions. We use machine learning methods to estimate how much each individual benefits from an intervention, and then reconcile its allocation with (i) the welfare weights assigned to different people; (ii) heterogeneous treatment effects of the intervention; and (iii) weights on different outcomes. We demonstrate this approach by analyzing Mexico's PROGRESA anti-poverty program. The analysis reveals that while the program prioritized certain subgroups -- such as indigenous households -- the fact that those groups benefited more implies that they were in fact assigned a lower welfare weight. The PROGRESA case illustrates how the method makes it possible to audit existing policies, and to design future policies that better align with values.

arXiv:2206.01092

Innovations in Integrating Machine Learning and Agent-Based Modeling of Biomedical Systems

Agent-based modeling (ABM) is a well-established paradigm for simulating complex systems via interactions between constituent entities. Machine learning (ML) refers to approaches whereby statistical algorithms 'learn' from data on their own, without imposing a priori theories of system behavior. Biological systems -- from molecules, to cells, to entire organisms -- consist of vast numbers of entities, governed by complex webs of interactions that span many spatiotemporal scales and exhibit nonlinearity, stochasticity and intricate coupling between entities. The macroscopic properties and collective dynamics of such systems are difficult to capture via continuum modelling and mean-field formalisms. ABM takes a 'bottom-up' approach that obviates these difficulties by enabling one to easily propose and test a set of well-defined 'rules' to be applied to the individual entities (agents) in a system. Evaluating a system and propagating its state over discrete time-steps effectively simulates the system, allowing observables to be computed and system properties to be analyzed. Because the rules that govern an ABM can be difficult to abstract and formulate from experimental data, there is an opportunity to use ML to help infer optimal, system-specific ABM rules. Once such rule-sets are devised, ABM calculations can generate a wealth of data, and ML can be applied there too -- e.g., to probe statistical measures that meaningfully describe a system's stochastic properties. As an example of synergy in the other direction (from ABM to ML), ABM simulations can generate realistic datasets for training ML algorithms (e.g., for regularization, to mitigate overfitting). In these ways, one can envision various synergistic ABM$\rightleftharpoons$ML loops. This review summarizes how ABM and ML have been integrated in contexts that span spatiotemporal scales, from cellular to population-level epidemiology.

arXiv:2206.01008공식 코드 · Papers with Code

Approximate Network Motif Mining Via Graph Learning

Frequent and structurally related subgraphs, also known as network motifs, are valuable features of many graph datasets. However, the high computational complexity of identifying motif sets in arbitrary datasets (motif mining) has limited their use in many real-world datasets. By automatically leveraging statistical properties of datasets, machine learning approaches have shown promise in several tasks with combinatorial complexity and are therefore a promising candidate for network motif mining. In this work we seek to facilitate the development of machine learning approaches aimed at motif mining. We propose a formulation of the motif mining problem as a node labelling task. In addition, we build benchmark datasets and evaluation metrics which test the ability of models to capture different aspects of motif discovery such as motif number, size, topology, and scarcity. Next, we propose MotiFiesta, a first attempt at solving this problem in a fully differentiable manner with promising results on challenging baselines. Finally, we demonstrate through MotiFiesta that this learning setting can be applied simultaneously to general-purpose data mining and interpretable feature extraction for graph classification tasks.

arXiv:2206.00885공식 코드 · Papers with Code

Coordinated Double Machine Learning

Double machine learning is a statistical method for leveraging complex black-box models to construct approximately unbiased treatment effect estimates given observational data with high-dimensional covariates, under the assumption of a partially linear model. The idea is to first fit on a subset of the samples two non-linear predictive models, one for the continuous outcome of interest and one for the observed treatment, and then to estimate a linear coefficient for the treatment using the remaining samples through a simple orthogonalized regression. While this methodology is flexible and can accommodate arbitrary predictive models, typically trained independently of one another, this paper argues that a carefully coordinated learning algorithm for deep neural networks may reduce the estimation bias. The improved empirical performance of the proposed method is demonstrated through numerical experiments on both simulated and real data.

arXiv:2206.00699

Studying the Practices of Deploying Machine Learning Projects on Docker

Docker is a containerization service that allows for convenient deployment of websites, databases, applications' APIs, and machine learning (ML) models with a few lines of code. Studies have recently explored the use of Docker for deploying general software projects with no specific focus on how Docker is used to deploy ML-based projects. In this study, we conducted an exploratory study to understand how Docker is being used to deploy ML-based projects. As the initial step, we examined the categories of ML-based projects that use Docker. We then examined why and how these projects use Docker, and the characteristics of the resulting Docker images. Our results indicate that six categories of ML-based projects use Docker for deployment, including ML Applications, MLOps/ AIOps, Toolkits, DL Frameworks, Models, and Documentation. We derived the taxonomy of 21 major categories representing the purposes of using Docker, including those specific to models such as model management tasks (e.g., testing, training). We then showed that ML engineers use Docker images mostly to help with the platform portability, such as transferring the software across the operating systems, runtimes such as GPU, and language constraints. However, we also found that more resources may be required to run the Docker images for building ML-based software projects due to the large number of files contained in the image layers with deeply nested directories. We hope to shed light on the emerging practices of deploying ML software projects using containers and highlight aspects that should be improved.

arXiv:2206.00667공식 코드 · Papers with Code

How Biased are Your Features?: Computing Fairness Influence Functions with Global Sensitivity Analysis

Fairness in machine learning has attained significant focus due to the widespread application in high-stake decision-making tasks. Unregulated machine learning classifiers can exhibit bias towards certain demographic groups in data, thus the quantification and mitigation of classifier bias is a central concern in fairness in machine learning. In this paper, we aim to quantify the influence of different features in a dataset on the bias of a classifier. To do this, we introduce the Fairness Influence Function (FIF). This function breaks down bias into its components among individual features and the intersection of multiple features. The key idea is to represent existing group fairness metrics as the difference of the scaled conditional variances in the classifier's prediction and apply a decomposition of variance according to global sensitivity analysis. To estimate FIFs, we instantiate an algorithm FairXplainer that applies variance decomposition of classifier's prediction following local regression. Experiments demonstrate that FairXplainer captures FIFs of individual feature and intersectional features, provides a better approximation of bias based on FIFs, demonstrates higher correlation of FIFs with fairness interventions, and detects changes in bias due to fairness affirmative/punitive actions in the classifier. The code is available at https://github.com/ReAILe/bias-explainer.

arXiv:2206.00574

Human Wellbeing and Machine Learning

There is a vast literature on the determinants of subjective wellbeing. International organisations and statistical offices are now collecting such survey data at scale. However, standard regression models explain surprisingly little of the variation in wellbeing, limiting our ability to predict it. In response, we here assess the potential of Machine Learning (ML) to help us better understand wellbeing. We analyse wellbeing data on over a million respondents from Germany, the UK, and the United States. In terms of predictive power, our ML approaches do perform better than traditional models. Although the size of the improvement is small in absolute terms, it turns out to be substantial when compared to that of key variables like health. We moreover find that drastically expanding the set of explanatory variables doubles the predictive power of both OLS and the ML approaches on unseen data. The variables identified as important by our ML algorithms - $i.e.$ material conditions, health, and meaningful social relations - are similar to those that have already been identified in the literature. In that sense, our data-driven ML results validate the findings from conventional approaches.