PyTorchKR Papers

최신 논문

총 8,922편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2010.06398

ProportionNet: Balancing Fairness and Revenue for Auction Design with Deep Learning

The design of revenue-maximizing auctions with strong incentive guarantees is a core concern of economic theory. Computational auctions enable online advertising, sourcing, spectrum allocation, and myriad financial markets. Analytic progress in this space is notoriously difficult; since Myerson's 1981 work characterizing single-item optimal auctions, there has been limited progress outside of restricted settings. A recent paper by D\"utting et al. circumvents analytic difficulties by applying deep learning techniques to, instead, approximate optimal auctions. In parallel, new research from Ilvento et al. and other groups has developed notions of fairness in the context of auction design. Inspired by these advances, in this paper, we extend techniques for approximating auctions using deep learning to address concerns of fairness while maintaining high revenue and strong incentive guarantees.

arXiv:2010.06192

Revisiting BFloat16 Training

State-of-the-art generic low-precision training algorithms use a mix of 16-bit and 32-bit precision, creating the folklore that 16-bit hardware compute units alone are not enough to maximize model accuracy. As a result, deep learning accelerators are forced to support both 16-bit and 32-bit floating-point units (FPUs), which is more costly than only using 16-bit FPUs for hardware design. We ask: can we train deep learning models only with 16-bit floating-point units, while still matching the model accuracy attained by 32-bit training? Towards this end, we study 16-bit-FPU training on the widely adopted BFloat16 unit. While these units conventionally use nearest rounding to cast output to 16-bit precision, we show that nearest rounding for model weight updates often cancels small updates, which degrades the convergence and model accuracy. Motivated by this, we study two simple techniques well-established in numerical analysis, stochastic rounding and Kahan summation, to remedy the model accuracy degradation in 16-bit-FPU training. We demonstrate that these two techniques can enable up to 7% absolute validation accuracy gain in 16-bit-FPU training. This leads to 0.1% lower to 0.2% higher validation accuracy compared to 32-bit training across seven deep learning applications.

arXiv:2010.06525

A Deep Learning Forecaster with Exogenous Variables for Day-Ahead Locational Marginal Price

Several approaches have been proposed to forecast day-ahead locational marginal price (daLMP) in deregulated energy markets. The rise of deep learning has motivated its use in energy price forecasts but most deep learning approaches fail to accommodate for exogenous variables, which have significant influence in the peaks and valleys of the daLMP. Accurate forecasts of the daLMP valleys are of crucial importance for power generators since one of the most important decisions they face is whether to sell power at a loss to prevent incurring in shutdown and start-up costs, or to bid at production cost and face the risk of shutting down. In this article we propose a deep learning model that incorporates both the history of daLMP and the effect of exogenous variables (e.g., forecasted load, weather data). A numerical study at the PJM independent system operator (ISO) illustrates how the proposed model outperforms traditional time series techniques while supporting risk-based analysis of shutdown decisions.

arXiv:2010.05774

Deep Learning for Information Systems Research

Artificial Intelligence (AI) has rapidly emerged as a key disruptive technology in the 21st century. At the heart of modern AI lies Deep Learning (DL), an emerging class of algorithms that has enabled today's platforms and organizations to operate at unprecedented efficiency, effectiveness, and scale. Despite significant interest, IS contributions in DL have been limited, which we argue is in part due to issues with defining, positioning, and conducting DL research. Recognizing the tremendous opportunity here for the IS community, this work clarifies, streamlines, and presents approaches for IS scholars to make timely and high-impact contributions. Related to this broader goal, this paper makes five timely contributions. First, we systematically summarize the major components of DL in a novel Deep Learning for Information Systems Research (DL-ISR) schematic that illustrates how technical DL processes are driven by key factors from an application environment. Second, we present a novel Knowledge Contribution Framework (KCF) to help IS scholars position their DL contributions for maximum impact. Third, we provide ten guidelines to help IS scholars generate rigorous and relevant DL-ISR in a systematic, high-quality fashion. Fourth, we present a review of prevailing journal and conference venues to examine how IS scholars have leveraged DL for various research inquiries. Finally, we provide a unique perspective on how IS scholars can formulate DL-ISR inquiries by carefully considering the interplay of business function(s), application areas(s), and the KCF. This perspective intentionally emphasizes inter-disciplinary, intra-disciplinary, and cross-IS tradition perspectives. Taken together, these contributions provide IS scholars a timely framework to advance the scale, scope, and impact of deep learning research.

arXiv:2010.05125

Learning Task-aware Robust Deep Learning Systems

Many works demonstrate that deep learning system is vulnerable to adversarial attack. A deep learning system consists of two parts: the deep learning task and the deep model. Nowadays, most existing works investigate the impact of the deep model on robustness of deep learning systems, ignoring the impact of the learning task. In this paper, we adopt the binary and interval label encoding strategy to redefine the classification task and design corresponding loss to improve robustness of the deep learning system. Our method can be viewed as improving the robustness of deep learning systems from both the learning task and deep model. Experimental results demonstrate that our learning task-aware method is much more robust than traditional classification while retaining the accuracy.

arXiv:2010.05496

Feature Extraction of Text for Deep Learning Algorithms: Application on Fake News Detection

Feature extraction is an important process of machine learning and deep learning, as the process make algorithms function more efficiently, and also accurate. In natural language processing used in deception detection such as fake news detection, several ways of feature extraction in statistical aspect had been introduced (e.g. N-gram). In this research, it will be shown that by using deep learning algorithms and alphabet frequencies of the original text of a news without any information about the sequence of the alphabet can actually be used to classify fake news and trustworthy ones in high accuracy (85\%). As this pre-processing method makes the data notably compact but also include the feature that is needed for the classifier, it seems that alphabet frequencies contains some useful features for understanding complex context or meaning of the original text.

논문 ID: 2020.10225

MH-COVIDNet: Diagnosis of COVID-19 using Deep Neural Networks and Meta-heuristic-based Feature Selection on X-ray Images

COVID-19 is a disease that causes symptoms in the lungs and causes deaths around the world. Studies are ongoing for the diagnosis and treatment of this disease, which is defined as a pandemic. Early diagnosis of this disease is important for human life. This process is progressing rapidly with diagnostic studies based on deep learning. Therefore, to contribute to this field, a deep learning-based approach that can be used for early diagnosis of the disease is proposed in our study. In this approach, a data set consisting of 3 classes of COVID19, normal and pneumonia lung X-ray images was created, with each class containing 364 images. Pre-processing was performed using the image contrast enhancement algorithm on the prepared data set and a new data set was obtained. Feature extraction was completed from this data set with deep learning models such as AlexNet, VGG19, GoogleNet, and ResNet. For the selection of the best potential features, two metaheuristic algorithms of binary particle swarm optimization and binary gray wolf optimization were used. After combining the features obtained in the feature selection of the enhancement data set, they were classified using SVM. The overall accuracy of the proposed approach was obtained as 99.38%. The results obtained by verification with two different metaheuristic algorithms proved that the approach we propose can help experts during COVID-19 diagnostic studies.

arXiv:2010.04413

A deep learning based interactive sketching system for fashion images design

In this work, we propose an interactive system to design diverse high-quality garment images from fashion sketches and the texture information. The major challenge behind this system is to generate high-quality and detailed texture according to the user-provided texture information. Prior works mainly use the texture patch representation and try to map a small texture patch to a whole garment image, hence unable to generate high-quality details. In contrast, inspired by intrinsic image decomposition, we decompose this task into texture synthesis and shading enhancement. In particular, we propose a novel bi-colored edge texture representation to synthesize textured garment images and a shading enhancer to render shading based on the grayscale edges. The bi-colored edge representation provides simple but effective texture cues and color constraints, so that the details can be better reconstructed. Moreover, with the rendered shading, the synthesized garment image becomes more vivid.

arXiv:2010.04365

DeepStreet: A deep learning powered urban street network generation module

In countries experiencing unprecedented waves of urbanization, there is a need for rapid and high quality urban street design. Our study presents a novel deep learning powered approach, DeepStreet (DS), for automatic street network generation that can be applied to the urban street design with local characteristics. DS is driven by a Convolutional Neural Network (CNN) that enables the interpolation of streets based on the areas of immediate vicinity. Specifically, the CNN is firstly trained to detect, recognize and capture the local features as well as the patterns of the existing street network sourced from the OpenStreetMap. With the trained CNN, DS is able to predict street networks' future expansion patterns within the predefined region conditioned on its surrounding street networks. To test the performance of DS, we apply it to an area in and around the Eixample area in the City of Barcelona, a well known example in the fields of urban and transport planning with iconic grid like street networks in the centre and irregular road alignments farther afield. The results show that DS can (1) detect and self cluster different types of complex street patterns in Barcelona; (2) predict both gridiron and irregular street and road networks. DS proves to have a great potential as a novel tool for designers to efficiently design the urban street network that well maintains the consistency across the existing and newly generated urban street network. Furthermore, the generated networks can serve as a benchmark to guide the local plan-making especially in rapidly developing cities.

arXiv:2010.04290

Deep Learning Meets Projective Clustering

A common approach for compressing NLP networks is to encode the embedding layer as a matrix $A\in\mathbb{R}^{n\times d}$, compute its rank-$j$ approximation $A_j$ via SVD, and then factor $A_j$ into a pair of matrices that correspond to smaller fully-connected layers to replace the original embedding layer. Geometrically, the rows of $A$ represent points in $\mathbb{R}^d$, and the rows of $A_j$ represent their projections onto the $j$-dimensional subspace that minimizes the sum of squared distances ("errors") to the points. In practice, these rows of $A$ may be spread around $k>1$ subspaces, so factoring $A$ based on a single subspace may lead to large errors that turn into large drops in accuracy. Inspired by \emph{projective clustering} from computational geometry, we suggest replacing this subspace by a set of $k$ subspaces, each of dimension $j$, that minimizes the sum of squared distances over every point (row in $A$) to its \emph{closest} subspace. Based on this approach, we provide a novel architecture that replaces the original embedding layer by a set of $k$ small layers that operate in parallel and are then recombined with a single fully-connected layer. Extensive experimental results on the GLUE benchmark yield networks that are both more accurate and smaller compared to the standard matrix factorization (SVD). For example, we further compress DistilBERT by reducing the size of the embedding layer by $40\%$ while incurring only a $0.5\%$ average drop in accuracy over all nine GLUE tasks, compared to a $2.8\%$ drop using the existing SVD approach. On RoBERTa we achieve $43\%$ compression of the embedding layer with less than a $0.8\%$ average drop in accuracy as compared to a $3\%$ drop previously. Open code for reproducing and extending our results is provided.

arXiv:2010.04589

Identifying Risk of Opioid Use Disorder for Patients Taking Opioid Medications with Deep Learning

The United States is experiencing an opioid epidemic, and there were more than 10 million opioid misusers aged 12 or older each year. Identifying patients at high risk of Opioid Use Disorder (OUD) can help to make early clinical interventions to reduce the risk of OUD. Our goal is to predict OUD patients among opioid prescription users through analyzing electronic health records with machine learning and deep learning methods. This will help us to better understand the diagnoses of OUD, providing new insights on opioid epidemic. Electronic health records of patients who have been prescribed with medications containing active opioid ingredients were extracted from Cerner Health Facts database between January 1, 2008 and December 31, 2017. Long Short-Term Memory (LSTM) models were applied to predict opioid use disorder risk in the future based on recent five encounters, and compared to Logistic Regression, Random Forest, Decision Tree and Dense Neural Network. Prediction performance was assessed using F-1 score, precision, recall, and AUROC. Our temporal deep learning model provided promising prediction results which outperformed other methods, with a F1 score of 0.8023 and AUCROC of 0.9369. The model can identify OUD related medications and vital signs as important features for the prediction. LSTM based temporal deep learning model is effective on predicting opioid use disorder using a patient past history of electronic health records, with minimal domain knowledge. It has potential to improve clinical decision support for early intervention and prevention to combat the opioid epidemic.

arXiv:2010.04556

Audio-Visual Speech Inpainting with Deep Learning

In this paper, we present a deep-learning-based framework for audio-visual speech inpainting, i.e., the task of restoring the missing parts of an acoustic speech signal from reliable audio context and uncorrupted visual information. Recent work focuses solely on audio-only methods and generally aims at inpainting music signals, which show highly different structure than speech. Instead, we inpaint speech signals with gaps ranging from 100 ms to 1600 ms to investigate the contribution that vision can provide for gaps of different duration. We also experiment with a multi-task learning approach where a phone recognition task is learned together with speech inpainting. Results show that the performance of audio-only speech inpainting approaches degrades rapidly when gaps get large, while the proposed audio-visual approach is able to plausibly restore missing information. In addition, we show that multi-task learning is effective, although the largest contribution to performance comes from vision.

arXiv:2010.04548

Deep Learning for Procedural Content Generation

Procedural content generation in video games has a long history. Existing procedural content generation methods, such as search-based, solver-based, rule-based and grammar-based methods have been applied to various content types such as levels, maps, character models, and textures. A research field centered on content generation in games has existed for more than a decade. More recently, deep learning has powered a remarkable range of inventions in content production, which are applicable to games. While some cutting-edge deep learning methods are applied on their own, others are applied in combination with more traditional methods, or in an interactive setting. This article surveys the various deep learning methods that have been applied to generate game content directly or indirectly, discusses deep learning methods that could be used for content generation purposes but are rarely used today, and envisages some limitations and potential future directions of deep learning for procedural content generation.

arXiv:2010.03951

MolDesigner: Interactive Design of Efficacious Drugs with Deep Learning

The efficacy of a drug depends on its binding affinity to the therapeutic target and pharmacokinetics. Deep learning (DL) has demonstrated remarkable progress in predicting drug efficacy. We develop MolDesigner, a human-in-the-loop web user-interface (UI), to assist drug developers leverage DL predictions to design more effective drugs. A developer can draw a drug molecule in the interface. In the backend, more than 17 state-of-the-art DL models generate predictions on important indices that are crucial for a drug's efficacy. Based on these predictions, drug developers can edit the drug molecule and reiterate until satisfaction. MolDesigner can make predictions in real-time with a latency of less than a second.

arXiv:2010.03639공식 코드 · Papers with Code

pymia: A Python package for data handling and evaluation in deep learning-based medical image analysis

Background and Objective: Deep learning enables tremendous progress in medical image analysis. One driving force of this progress are open-source frameworks like TensorFlow and PyTorch. However, these frameworks rarely address issues specific to the domain of medical image analysis, such as 3-D data handling and distance metrics for evaluation. pymia, an open-source Python package, tries to address these issues by providing flexible data handling and evaluation independent of the deep learning framework. Methods: The pymia package provides data handling and evaluation functionalities. The data handling allows flexible medical image handling in every commonly used format (e.g., 2-D, 2.5-D, and 3-D; full- or patch-wise). Even data beyond images like demographics or clinical reports can easily be integrated into deep learning pipelines. The evaluation allows stand-alone result calculation and reporting, as well as performance monitoring during training using a vast amount of domain-specific metrics for segmentation, reconstruction, and regression. Results: The pymia package is highly flexible, allows for fast prototyping, and reduces the burden of implementing data handling routines and evaluation methods. While data handling and evaluation are independent of the deep learning framework used, they can easily be integrated into TensorFlow and PyTorch pipelines. The developed package was successfully used in a variety of research projects for segmentation, reconstruction, and regression. Conclusions: The pymia package fills the gap of current deep learning frameworks regarding data handling and evaluation in medical image analysis. It is available at https://github.com/rundherum/pymia and can directly be installed from the Python Package Index using pip install pymia.

arXiv:2010.03902

IRX-1D: A Simple Deep Learning Architecture for Remote Sensing Classifications

We proposes a simple deep learning architecture combining elements of Inception, ResNet and Xception networks. Four new datasets were used for classification with both small and large training samples. Results in terms of classification accuracy suggests improved performance by proposed architecture in comparison to Bayesian optimised 2D-CNN with small training samples. Comparison of results using small training sample with Indiana Pines hyperspectral dataset suggests comparable or better performance by proposed architecture than nine reported works using different deep learning architectures. In spite of achieving high classification accuracy with limited training samples, comparison of classified image suggests different land cover classes are assigned to same area when compared with the classified image provided by the model trained using large training samples with all datasets.

arXiv:2010.03701

Differentially Private Deep Learning with Direct Feedback Alignment

Standard methods for differentially private training of deep neural networks replace back-propagated mini-batch gradients with biased and noisy approximations to the gradient. These modifications to training often result in a privacy-preserving model that is significantly less accurate than its non-private counterpart. We hypothesize that alternative training algorithms may be more amenable to differential privacy. Specifically, we examine the suitability of direct feedback alignment (DFA). We propose the first differentially private method for training deep neural networks with DFA and show that it achieves significant gains in accuracy (often by 10-20%) compared to backprop-based differentially private training on a variety of architectures (fully connected, convolutional) and datasets.

arXiv:2010.04062

MIA-Prognosis: A Deep Learning Framework to Predict Therapy Response

Predicting clinical outcome is remarkably important but challenging. Research efforts have been paid on seeking significant biomarkers associated with the therapy response or/and patient survival. However, these biomarkers are generally costly and invasive, and possibly dissatifactory for novel therapy. On the other hand, multi-modal, heterogeneous, unaligned temporal data is continuously generated in clinical practice. This paper aims at a unified deep learning approach to predict patient prognosis and therapy response, with easily accessible data, e.g., radiographics, laboratory and clinical information. Prior arts focus on modeling single data modality, or ignore the temporal changes. Importantly, the clinical time series is asynchronous in practice, i.e., recorded with irregular intervals. In this study, we formalize the prognosis modeling as a multi-modal asynchronous time series classification task, and propose a MIA-Prognosis framework with Measurement, Intervention and Assessment (MIA) information to predict therapy response, where a Simple Temporal Attention (SimTA) module is developed to process the asynchronous time series. Experiments on synthetic dataset validate the superiory of SimTA over standard RNN-based approaches. Furthermore, we experiment the proposed method on an in-house, retrospective dataset of real-world non-small cell lung cancer patients under anti-PD-1 immunotherapy. The proposed method achieves promising performance on predicting the immunotherapy response. Notably, our predictive model could further stratify low-risk and high-risk patients in terms of long-term survival.

arXiv:2010.02727

Symbolic Techniques for Deep Learning: Challenges and Opportunities

As the number of deep learning frameworks increase and certain ones gain popularity, it spurs the discussion of what methodologies are employed by these frameworks and the reasoning behind them. The goal of this survey is to study how symbolic techniques are utilized in deep learning. To do this, we look at some of the most popular deep learning frameworks being used today, including TensorFlow, Keras, PyTorch, and MXNet. While these frameworks greatly differ from one another, many of them use symbolic techniques, whether it be symbolic execution, graphs, or programming. We focus this paper on symbolic techniques because they influence not only how neural networks are built but also the way in which they are executed. Limitations of symbolic techniques have led to efforts in integrating symbolic and nonsymbolic aspects in deep learning, opening up new possibilities for symbolic techniques. For example, the Gluon API by Apache MXNet bridges the gap between imperative programming and symbolic execution through hybridization. Frameworks such as JANUS attempt to translate imperative programs into symbolic graphs, while approaches like DeepCheck attempt to use symbolic execution to analyze and validate imperative neural network programs. Symbolic analysis has also been paired with concrete execution in a technique called concolic testing in order to better test deep neural networks. Our study of these developments exemplifies just a few of the many ways the symbolic techniques employed by popular frameworks have the opportunity to be altered and utilized to achieve better performance.