PyTorchKR Papers

최신 논문

총 2,063편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2310.19109공식 코드 · Papers with Code

Dynamic Task and Weight Prioritization Curriculum Learning for Multimodal Imagery

This paper explores post-disaster analytics using multimodal deep learning models trained with curriculum learning method. Studying post-disaster analytics is important as it plays a crucial role in mitigating the impact of disasters by providing timely and accurate insights into the extent of damage and the allocation of resources. We propose a curriculum learning strategy to enhance the performance of multimodal deep learning models. Curriculum learning emulates the progressive learning sequence in human education by training deep learning models on increasingly complex data. Our primary objective is to develop a curriculum-trained multimodal deep learning model, with a particular focus on visual question answering (VQA) capable of jointly processing image and text data, in conjunction with semantic segmentation for disaster analytics using the FloodNet\footnote{https://github.com/BinaLab/FloodNet-Challenge-EARTHVISION2021} dataset. To achieve this, U-Net model is used for semantic segmentation and image encoding. A custom built text classifier is used for visual question answering. Existing curriculum learning methods rely on manually defined difficulty functions. We introduce a novel curriculum learning approach termed Dynamic Task and Weight Prioritization (DATWEP), which leverages a gradient-based method to automatically decide task difficulty during curriculum learning training, thereby eliminating the need for explicit difficulty computation. The integration of DATWEP into our multimodal model shows improvement on VQA performance. Source code is available at https://github.com/fualsan/DATWEP.

arXiv:2310.18636공식 코드 · Papers with Code

Electrical Impedance Tomography: A Fair Comparative Study on Deep Learning and Analytic-based Approaches

Electrical Impedance Tomography (EIT) is a powerful imaging technique with diverse applications, e.g., medical diagnosis, industrial monitoring, and environmental studies. The EIT inverse problem is about inferring the internal conductivity distribution of an object from measurements taken on its boundary. It is severely ill-posed, necessitating advanced computational methods for accurate image reconstructions. Recent years have witnessed significant progress, driven by innovations in analytic-based approaches and deep learning. This review explores techniques for solving the EIT inverse problem, focusing on the interplay between contemporary deep learning-based strategies and classical analytic-based methods. Four state-of-the-art deep learning algorithms are rigorously examined, harnessing the representational capabilities of deep neural networks to reconstruct intricate conductivity distributions. In parallel, two analytic-based methods, rooted in mathematical formulations and regularisation techniques, are dissected for their strengths and limitations. These methodologies are evaluated through various numerical experiments, encompassing diverse scenarios that reflect real-world complexities. A suite of performance metrics is employed to assess the efficacy of these methods. These metrics collectively provide a nuanced understanding of the methods' ability to capture essential features and delineate complex conductivity patterns. One novel feature of the study is the incorporation of variable conductivity scenarios, introducing a level of heterogeneity that mimics textured inclusions. This departure from uniform conductivity assumptions mimics realistic scenarios where tissues or materials exhibit spatially varying electrical properties. Exploring how each method responds to such variable conductivity scenarios opens avenues for understanding their robustness and adaptability.

arXiv:2310.17126공식 코드 · Papers with Code

Deep Learning on SAR Imagery: Transfer Learning Versus Randomly Initialized Weights

Deploying deep learning on Synthetic Aperture Radar (SAR) data is becoming more common for mapping purposes. One such case is sea ice, which is highly dynamic and rapidly changes as a result of the combined effect of wind, temperature, and ocean currents. Therefore, frequent mapping of sea ice is necessary to ensure safe marine navigation. However, there is a general shortage of expert-labeled data to train deep learning algorithms. Fine-tuning a pre-trained model on SAR imagery is a potential solution. In this paper, we compare the performance of deep learning models trained from scratch using randomly initialized weights against pre-trained models that we fine-tune for this purpose. Our results show that pre-trained models lead to better results, especially on test samples from the melt season.

arXiv:2310.17010공식 코드 · Papers with Code

This Reads Like That: Deep Learning for Interpretable Natural Language Processing

Prototype learning, a popular machine learning method designed for inherently interpretable decisions, leverages similarities to learned prototypes for classifying new data. While it is mainly applied in computer vision, in this work, we build upon prior research and further explore the extension of prototypical networks to natural language processing. We introduce a learned weighted similarity measure that enhances the similarity computation by focusing on informative dimensions of pre-trained sentence embeddings. Additionally, we propose a post-hoc explainability mechanism that extracts prediction-relevant words from both the prototype and input sentences. Finally, we empirically demonstrate that our proposed method not only improves predictive performance on the AG News and RT Polarity datasets over a previous prototype-based approach, but also improves the faithfulness of explanations compared to rationale-based recurrent convolutions.

arXiv:2310.16499공식 코드 · Papers with Code

Data Optimization in Deep Learning: A Survey

Large-scale, high-quality data are considered an essential factor for the successful application of many deep learning techniques. Meanwhile, numerous real-world deep learning tasks still have to contend with the lack of sufficient amounts of high-quality data. Additionally, issues such as model robustness, fairness, and trustworthiness are also closely related to training data. Consequently, a huge number of studies in the existing literature have focused on the data aspect in deep learning tasks. Some typical data optimization techniques include data augmentation, logit perturbation, sample weighting, and data condensation. These techniques usually come from different deep learning divisions and their theoretical inspirations or heuristic motivations may seem unrelated to each other. This study aims to organize a wide range of existing data optimization methodologies for deep learning from the previous literature, and makes the effort to construct a comprehensive taxonomy for them. The constructed taxonomy considers the diversity of split dimensions, and deep sub-taxonomies are constructed for each dimension. On the basis of the taxonomy, connections among the extensive data optimization methods for deep learning are built in terms of four aspects. We probe into rendering several promising and interesting future directions. The constructed taxonomy and the revealed connections will enlighten the better understanding of existing methods and the design of novel data optimization techniques. Furthermore, our aspiration for this survey is to promote data optimization as an independent subdivision of deep learning. A curated, up-to-date list of resources related to data optimization in deep learning is available at \url{https://github.com/YaoRujing/Data-Optimization}.

arXiv:2310.16485공식 코드 · Papers with Code

A Comprehensive Python Library for Deep Learning-Based Event Detection in Multivariate Time Series Data and Information Retrieval in NLP

Event detection in time series data is crucial in various domains, including finance, healthcare, cybersecurity, and science. Accurately identifying events in time series data is vital for making informed decisions, detecting anomalies, and predicting future trends. Despite extensive research exploring diverse methods for event detection in time series, with deep learning approaches being among the most advanced, there is still room for improvement and innovation in this field. In this paper, we present a new deep learning supervised method for detecting events in multivariate time series data. Our method combines four distinct novelties compared to existing deep-learning supervised methods. Firstly, it is based on regression instead of binary classification. Secondly, it does not require labeled datasets where each point is labeled; instead, it only requires reference events defined as time points or intervals of time. Thirdly, it is designed to be robust by using a stacked ensemble learning meta-model that combines deep learning models, ranging from classic feed-forward neural networks (FFNs) to state-of-the-art architectures like transformers. This ensemble approach can mitigate individual model weaknesses and biases, resulting in more robust predictions. Finally, to facilitate practical implementation, we have developed a Python package to accompany our proposed method. The package, called eventdetector-ts, can be installed through the Python Package Index (PyPI). In this paper, we present our method and provide a comprehensive guide on the usage of the package. We showcase its versatility and effectiveness through different real-world use cases from natural language processing (NLP) to financial security domains.

arXiv:2310.16210공식 코드 · Papers with Code

Semantic Segmentation in Satellite Hyperspectral Imagery by Deep Learning

Satellites are increasingly adopting on-board AI to optimize operations and increase autonomy through in-orbit inference. The use of Deep Learning (DL) models for segmentation in hyperspectral imagery offers advantages for remote sensing applications. In this work, we train and test 20 models for multi-class segmentation in hyperspectral imagery, selected for their potential in future space deployment. These models include 1D and 2D Convolutional Neural Networks (CNNs) and the latest vision transformers (ViTs). We propose a lightweight 1D-CNN model, 1D-Justo-LiuNet, which outperforms state-of-the-art models in the hypespectral domain. 1D-Justo-LiuNet exceeds the performance of 2D-CNN UNets and outperforms Apple's lightweight vision transformers designed for mobile inference. 1D-Justo-LiuNet achieves the highest accuracy (0.93) with the smallest model size (4,563 parameters) among all tested models, while maintaining fast inference. Unlike 2D-CNNs and ViTs, which encode both spectral and spatial information, 1D-Justo-LiuNet focuses solely on the rich spectral features in hyperspectral data, benefitting from the high-dimensional feature space. Our findings are validated across various satellite datasets, with the HYPSO-1 mission serving as the primary case study for sea, land, and cloud segmentation. We further confirm our conclusions through generalization tests on other hyperspectral missions, such as NASA's EO-1. Based on its superior performance and compact size, we conclude that 1D-Justo-LiuNet is highly suitable for in-orbit deployment, providing an effective solution for optimizing and automating satellite operations at edge.

arXiv:2310.14423공식 코드 · Papers with Code

A Quadratic Synchronization Rule for Distributed Deep Learning

In distributed deep learning with data parallelism, synchronizing gradients at each training step can cause a huge communication overhead, especially when many nodes work together to train large models. Local gradient methods, such as Local SGD, address this issue by allowing workers to compute locally for $H$ steps without synchronizing with others, hence reducing communication frequency. While $H$ has been viewed as a hyperparameter to trade optimization efficiency for communication cost, recent research indicates that setting a proper $H$ value can lead to generalization improvement. Yet, selecting a proper $H$ is elusive. This work proposes a theory-grounded method for determining $H$, named the Quadratic Synchronization Rule (QSR), which recommends dynamically setting $H$ in proportion to $\frac{1}{\eta^2}$ as the learning rate $\eta$ decays over time. Extensive ImageNet experiments on ResNet and ViT show that local gradient methods with QSR consistently improve the test accuracy over other synchronization strategies. Compared with the standard data parallel training, QSR enables Local AdamW on ViT-B to cut the training time on 16 or 64 GPUs down from 26.7 to 20.2 hours or from 8.6 to 5.5 hours and, at the same time, achieves $1.16\%$ or $0.84\%$ higher top-1 validation accuracy.

arXiv:2310.14743공식 코드 · Papers with Code

The Safety Challenges of Deep Learning in Real-World Type 1 Diabetes Management

Blood glucose simulation allows the effectiveness of type 1 diabetes (T1D) management strategies to be evaluated without patient harm. Deep learning algorithms provide a promising avenue for extending simulator capabilities; however, these algorithms are limited in that they do not necessarily learn physiologically correct glucose dynamics and can learn incorrect and potentially dangerous relationships from confounders in training data. This is likely to be more important in real-world scenarios, as data is not collected under strict research protocol. This work explores the implications of using deep learning algorithms trained on real-world data to model glucose dynamics. Free-living data was processed from the OpenAPS Data Commons and supplemented with patient-reported tags of challenging diabetes events, constituting one of the most detailed real-world T1D datasets. This dataset was used to train and evaluate state-of-the-art glucose simulators, comparing their prediction error across safety critical scenarios and assessing the physiological appropriateness of the learned dynamics using Shapley Additive Explanations (SHAP). While deep learning prediction accuracy surpassed the widely-used mathematical simulator approach, the model deteriorated in safety critical scenarios and struggled to leverage self-reported meal and exercise information. SHAP value analysis also indicated the model had fundamentally confused the roles of insulin and carbohydrates, which is one of the most basic T1D management principles. This work highlights the importance of considering physiological appropriateness when using deep learning to model real-world systems in T1D and healthcare more broadly, and provides recommendations for building models that are robust to real-world data constraints.

arXiv:2310.13369공식 코드 · Papers with Code

SigFormer: Signature Transformers for Deep Hedging

Deep hedging is a promising direction in quantitative finance, incorporating models and techniques from deep learning research. While giving excellent hedging strategies, models inherently requires careful treatment in designing architectures for neural networks. To mitigate such difficulties, we introduce SigFormer, a novel deep learning model that combines the power of path signatures and transformers to handle sequential data, particularly in cases with irregularities. Path signatures effectively capture complex data patterns, while transformers provide superior sequential attention. Our proposed model is empirically compared to existing methods on synthetic data, showcasing faster learning and enhanced robustness, especially in the presence of irregular underlying price data. Additionally, we validate our model performance through a real-world backtest on hedging the SP 500 index, demonstrating positive outcomes.

arXiv:2310.13349공식 코드 · Papers with Code

DeepFDR: A Deep Learning-based False Discovery Rate Control Method for Neuroimaging Data

Voxel-based multiple testing is widely used in neuroimaging data analysis. Traditional false discovery rate (FDR) control methods often ignore the spatial dependence among the voxel-based tests and thus suffer from substantial loss of testing power. While recent spatial FDR control methods have emerged, their validity and optimality remain questionable when handling the complex spatial dependencies of the brain. Concurrently, deep learning methods have revolutionized image segmentation, a task closely related to voxel-based multiple testing. In this paper, we propose DeepFDR, a novel spatial FDR control method that leverages unsupervised deep learning-based image segmentation to address the voxel-based multiple testing problem. Numerical studies, including comprehensive simulations and Alzheimer's disease FDG-PET image analysis, demonstrate DeepFDR's superiority over existing methods. DeepFDR not only excels in FDR control and effectively diminishes the false nondiscovery rate, but also boasts exceptional computational efficiency highly suited for tackling large-scale neuroimaging data.

arXiv:2310.12802공식 코드 · Papers with Code

An effective theory of collective deep learning

Unraveling the emergence of collective learning in systems of coupled artificial neural networks points to broader implications for machine learning, neuroscience, and society. Here we introduce a minimal model that condenses several recent decentralized algorithms by considering a competition between two terms: the local learning dynamics in the parameters of each neural network unit, and a diffusive coupling among units that tends to homogenize the parameters of the ensemble. We derive an effective theory for linear networks to show that the coarse-grained behavior of our system is equivalent to a deformed Ginzburg-Landau model with quenched disorder. This framework predicts depth-dependent disorder-order-disorder phase transitions in the parameters' solutions that reveal a depth-delayed onset of a collective learning phase and a low-rank microscopic learning path. We validate the theory in coupled ensembles of realistic neural networks trained on the MNIST dataset under privacy constraints. Interestingly, experiments confirm that individual networks -- trained on private data -- can fully generalize to unseen data classes when the collective learning phase emerges. Our work establishes the physics of collective learning and contributes to the mechanistic interpretability of deep learning in decentralized settings.

arXiv:2310.09789공식 코드 · Papers with Code

FLrce: Resource-Efficient Federated Learning with Early-Stopping Strategy

Federated Learning (FL) achieves great popularity in the Internet of Things (IoT) as a powerful interface to offer intelligent services to customers while maintaining data privacy. Under the orchestration of a server, edge devices (also called clients in FL) collaboratively train a global deep-learning model without sharing any local data. Nevertheless, the unequal training contributions among clients have made FL vulnerable, as clients with heavily biased datasets can easily compromise FL by sending malicious or heavily biased parameter updates. Furthermore, the resource shortage issue of the network also becomes a bottleneck. Due to overwhelming computation overheads generated by training deep-learning models on edge devices, and significant communication overheads for transmitting deep-learning models across the network, enormous amounts of resources are consumed in the FL process. This encompasses computation resources like energy and communication resources like bandwidth. To comprehensively address these challenges, in this paper, we present FLrce, an efficient FL framework with a relationship-based client selection and early-stopping strategy. FLrce accelerates the FL process by selecting clients with more significant effects, enabling the global model to converge to a high accuracy in fewer rounds. FLrce also leverages an early stopping mechanism that terminates FL in advance to save communication and computation resources. Experiment results show that, compared with existing efficient FL frameworks, FLrce improves the computation and communication efficiency by at least 30% and 43% respectively.

arXiv:2310.09443공식 코드 · Papers with Code

G10: Enabling An Efficient Unified GPU Memory and Storage Architecture with Smart Tensor Migrations

To break the GPU memory wall for scaling deep learning workloads, a variety of architecture and system techniques have been proposed recently. Their typical approaches include memory extension with flash memory and direct storage access. However, these techniques still suffer from suboptimal performance and introduce complexity to the GPU memory management, making them hard to meet the scalability requirement of deep learning workloads today. In this paper, we present a unified GPU memory and storage architecture named G10 driven by the fact that the tensor behaviors of deep learning workloads are highly predictable. G10 integrates the host memory, GPU memory, and flash memory into a unified memory space, to scale the GPU memory capacity while enabling transparent data migrations. Based on this unified GPU memory and storage architecture, G10 utilizes compiler techniques to characterize the tensor behaviors in deep learning workloads. Therefore, it can schedule data migrations in advance by considering the available bandwidth of flash memory and host memory. The cooperative mechanism between deep learning compilers and the unified memory architecture enables G10 to hide data transfer overheads in a transparent manner. We implement G10 based on an open-source GPU simulator. Our experiments demonstrate that G10 outperforms state-of-the-art GPU memory solutions by up to 1.75$\times$, without code modifications to deep learning workloads. With the smart data migration mechanism, G10 can reach 90.3\% of the performance of the ideal case assuming unlimited GPU memory.

arXiv:2310.10550공식 코드 · Papers with Code

Deep learning applied to EEG data with different montages using spatial attention

The ability of Deep Learning to process and extract relevant information in complex brain dynamics from raw EEG data has been demonstrated in various recent works. Deep learning models, however, have also been shown to perform best on large corpora of data. When processing EEG, a natural approach is to combine EEG datasets from different experiments to train large deep-learning models. However, most EEG experiments use custom channel montages, requiring the data to be transformed into a common space. Previous methods have used the raw EEG signal to extract features of interest and focused on using a common feature space across EEG datasets. While this is a sensible approach, it underexploits the potential richness of EEG raw data. Here, we explore using spatial attention applied to EEG electrode coordinates to perform channel harmonization of raw EEG data, allowing us to train deep learning on EEG data using different montages. We test this model on a gender classification task. We first show that spatial attention increases model performance. Then, we show that a deep learning model trained on data using different channel montages performs significantly better than deep learning models trained on fixed 23- and 128-channel data montages.

arXiv:2310.10537공식 코드 · Papers with Code

Microscaling Data Formats for Deep Learning

Narrow bit-width data formats are key to reducing the computational and storage costs of modern deep learning applications. This paper evaluates Microscaling (MX) data formats that combine a per-block scaling factor with narrow floating-point and integer types for individual elements. MX formats balance the competing needs of hardware efficiency, model accuracy, and user friction. Empirical results on over two dozen benchmarks demonstrate practicality of MX data formats as a drop-in replacement for baseline FP32 for AI inference and training with low user friction. We also show the first instance of training generative language models at sub-8-bit weights, activations, and gradients with minimal accuracy loss and no modifications to the training recipe.

arXiv:2310.08717공식 코드 · Papers with Code

Designing Observables for Measurements with Deep Learning

Many analyses in particle and nuclear physics use simulations to infer fundamental, effective, or phenomenological parameters of the underlying physics models. When the inference is performed with unfolded cross sections, the observables are designed using physics intuition and heuristics. We propose to design targeted observables with machine learning. Unfolded, differential cross sections in a neural network output contain the most information about parameters of interest and can be well-measured by construction. The networks are trained using a custom loss function that rewards outputs that are sensitive to the parameter(s) of interest while simultaneously penalizing outputs that are different between particle-level and detector-level (to minimize detector distortions). We demonstrate this idea in simulation using two physics models for inclusive measurements in deep inelastic scattering. We find that the new approach is more sensitive than classical observables at distinguishing the two models and also has a reduced unfolding uncertainty due to the reduced detector distortions.

arXiv:2310.09203공식 코드 · Papers with Code

SiamAF: Learning Shared Information from ECG and PPG Signals for Robust Atrial Fibrillation Detection

Atrial fibrillation (AF) is the most common type of cardiac arrhythmia. It is associated with an increased risk of stroke, heart failure, and other cardiovascular complications, but can be clinically silent. Passive AF monitoring with wearables may help reduce adverse clinical outcomes related to AF. Detecting AF in noisy wearable data poses a significant challenge, leading to the emergence of various deep learning techniques. Previous deep learning models learn from a single modality, either electrocardiogram (ECG) or photoplethysmography (PPG) signals. However, deep learning models often struggle to learn generalizable features and rely on features that are more susceptible to corruption from noise, leading to sub-optimal performances in certain scenarios, especially with low-quality signals. Given the increasing availability of ECG and PPG signal pairs from wearables and bedside monitors, we propose a new approach, SiamAF, leveraging a novel Siamese network architecture and joint learning loss function to learn shared information from both ECG and PPG signals. At inference time, the proposed model is able to predict AF from either PPG or ECG and outperforms baseline methods on three external test sets. It learns medically relevant features as a result of our novel architecture design. The proposed model also achieves comparable performance to traditional learning regimes while requiring much fewer training labels, providing a potential approach to reduce future reliance on manual labeling.

arXiv:2310.04415공식 코드 · Papers with Code

Why Do We Need Weight Decay in Modern Deep Learning?

Weight decay is a broadly used technique for training state-of-the-art deep networks from image classification to large language models. Despite its widespread usage and being extensively studied in the classical literature, its role remains poorly understood for deep learning. In this work, we highlight that the role of weight decay in modern deep learning is different from its regularization effect studied in classical learning theory. For deep networks on vision tasks trained with multipass SGD, we show how weight decay modifies the optimization dynamics enhancing the ever-present implicit regularization of SGD via the loss stabilization mechanism. In contrast, for large language models trained with nearly one-epoch training, we describe how weight decay balances the bias-variance tradeoff in stochastic optimization leading to lower training loss and improved training stability. Overall, we present a unifying perspective from ResNets on vision tasks to LLMs: weight decay is never useful as an explicit regularizer but instead changes the training dynamics in a desirable way. The code is available at https://github.com/tml-epfl/why-weight-decay

arXiv:2310.04345공식 코드 · Papers with Code

Deep Learning for Two-Stage Robust Integer Optimization

Robust optimization is an established framework for modeling optimization problems with uncertain parameters. While static robust optimization is often criticized for being too conservative, two-stage (or adjustable) robust optimization (2RO) provides a less conservative alternative by allowing some decisions to be made after the uncertain parameters have been revealed. Unfortunately, in the case of integer decision variables, existing solution methods for 2RO typically fail to solve large-scale instances, limiting the applicability of this modeling paradigm to simple cases. We propose Neur2RO, a deep-learning-augmented instantiation of the column-and-constraint-generation (CCG) algorithm, which expands the applicability of the 2RO framework to large-scale instances with integer decisions in both stages. A custom-designed neural network is trained to estimate the optimal value and feasibility of the second-stage problem. The network can be incorporated into CCG, leading to more computationally tractable subproblems in each of its iterations. The resulting algorithm enjoys approximation guarantees which depend on the neural network's prediction error. In our experiments, Neur2RO produces high-quality solutions quickly, outperforming state-of-the-art methods on two-stage knapsack, capital budgeting, and facility location problems. Compared to existing methods, which often run for hours, Neur2RO finds better solutions in a few seconds or minutes. Our code is available at https://github.com/khalil-research/Neur2RO.