PyTorchKR Papers

최신 논문

총 12,307편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:1912.00672

On Classifying Sepsis Heterogeneity in the ICU: Insight Using Machine Learning

Current machine learning models aiming to predict sepsis from Electronic Health Records (EHR) do not account for the heterogeneity of the condition, despite its emerging importance in prognosis and treatment. This work demonstrates the added value of stratifying the types of organ dysfunction observed in patients who develop sepsis in the ICU in improving the ability to recognise patients at risk of sepsis from their EHR data. Using an ICU dataset of 13,728 records, we identify clinically significant sepsis subpopulations with distinct organ dysfunction patterns. Classification experiments using Random Forest, Gradient Boost Trees and Support Vector Machines, aiming to distinguish patients who develop sepsis in the ICU from those who do not, show that features selected using sepsis subpopulations as background knowledge yield a superior performance regardless of the classification model used. Our findings can steer machine learning efforts towards more personalised models for complex conditions including sepsis.

arXiv:1912.00622공식 코드 · Papers with Code

Patchy Image Structure Classification Using Multi-Orientation Region Transform

Exterior contour and interior structure are both vital features for classifying objects. However, most of the existing methods consider exterior contour feature and internal structure feature separately, and thus fail to function when classifying patchy image structures that have similar contours and flexible structures. To address above limitations, this paper proposes a novel Multi-Orientation Region Transform (MORT), which can effectively characterize both contour and structure features simultaneously, for patchy image structure classification. MORT is performed over multiple orientation regions at multiple scales to effectively integrate patchy features, and thus enables a better description of the shape in a coarse-to-fine manner. Moreover, the proposed MORT can be extended to combine with the deep convolutional neural network techniques, for further enhancement of classification accuracy. Very encouraging experimental results on the challenging ultra-fine-grained cultivar recognition task, insect wing recognition task, and large variation butterfly recognition task are obtained, which demonstrate the effectiveness and superiority of the proposed MORT over the state-of-the-art methods in classifying patchy image structures. Our code and three patchy image structure datasets are available at: https://github.com/XiaohanYu-GU/MReT2019.

arXiv:1912.00412

MetAdapt: Meta-Learned Task-Adaptive Architecture for Few-Shot Classification

Few-Shot Learning (FSL) is a topic of rapidly growing interest. Typically, in FSL a model is trained on a dataset consisting of many small tasks (meta-tasks) and learns to adapt to novel tasks that it will encounter during test time. This is also referred to as meta-learning. Another topic closely related to meta-learning with a lot of interest in the community is Neural Architecture Search (NAS), automatically finding optimal architecture instead of engineering it manually. In this work, we combine these two aspects of meta-learning. So far, meta-learning FSL methods have focused on optimizing parameters of pre-defined network architectures, in order to make them easily adaptable to novel tasks. Moreover, it was observed that, in general, larger architectures perform better than smaller ones up to a certain saturation point (where they start to degrade due to over-fitting). However, little attention has been given to explicitly optimizing the architectures for FSL, nor to an adaptation of the architecture at test time to particular novel tasks. In this work, we propose to employ tools inspired by the Differentiable Neural Architecture Search (D-NAS) literature in order to optimize the architecture for FSL without over-fitting. Additionally, to make the architecture task adaptive, we propose the concept of `MetAdapt Controller' modules. These modules are added to the model and are meta-trained to predict the optimal network connections for a given novel task. Using the proposed approach we observe state-of-the-art results on two popular few-shot benchmarks: miniImageNet and FC100.

arXiv:1912.00905

Matrix sketching for supervised classification with imbalanced classes

Matrix sketching is a recently developed data compression technique. An input matrix A is efficiently approximated with a smaller matrix B, so that B preserves most of the properties of A up to some guaranteed approximation ratio. In so doing numerical operations on big data sets become faster. Sketching algorithms generally use random projections to compress the original dataset and this stochastic generation process makes them amenable to statistical analysis. The statistical properties of sketching algorithms have been widely studied in the context of multiple linear regression. In this paper we propose matrix sketching as a tool for rebalancing class sizes in supervised classification with imbalanced classes. It is well-known in fact that class imbalance may lead to poor classification performances especially as far as the minority class is concerned.

arXiv:1912.00552공식 코드 · Papers with Code

Sparse Graph Attention Networks

Graph Neural Networks (GNNs) have proved to be an effective representation learning framework for graph-structured data, and have achieved state-of-the-art performance on many practical predictive tasks, such as node classification, link prediction and graph classification. Among the variants of GNNs, Graph Attention Networks (GATs) learn to assign dense attention coefficients over all neighbors of a node for feature aggregation, and improve the performance of many graph learning tasks. However, real-world graphs are often very large and noisy, and GATs are prone to overfitting if not regularized properly. Even worse, the local aggregation mechanism of GATs may fail on disassortative graphs, where nodes within local neighborhood provide more noise than useful information for feature aggregation. In this paper, we propose Sparse Graph Attention Networks (SGATs) that learn sparse attention coefficients under an $L_0$-norm regularization, and the learned sparse attentions are then used for all GNN layers, resulting in an edge-sparsified graph. By doing so, we can identify noisy/task-irrelevant edges, and thus perform feature aggregation on most informative neighbors. Extensive experiments on synthetic and real-world graph learning benchmarks demonstrate the superior performance of SGATs. In particular, SGATs can remove about 50\%-80\% edges from large assortative graphs, while retaining similar classification accuracies. On disassortative graphs, SGATs prune majority of noisy edges and outperform GATs in classification accuracies by significant margins. Furthermore, the removed edges can be interpreted intuitively and quantitatively. To the best of our knowledge, this is the first graph learning algorithm that shows significant redundancies in graphs and edge-sparsified graphs can achieve similar or sometimes higher predictive performances than original graphs.

arXiv:1912.00519

Location Forensics of Media Recordings Utilizing Cascaded SVM and Pole-matching Classifiers

Information regarding the location of power distribution grid can be extracted from the power signature embedded in the multimedia signals (e.g., audio, video data) recorded near electrical activities. This implicit mechanism of identifying the origin-of-recording can be a very promising tool for multimedia forensics and security applications. In this work, we have developed a novel grid-of-origin identification system from media recording that consists of a number of support vector machine (SVM) followed by pole-matching (PM) classifiers. First, we determine the nominal frequency of the grid (50 or 60 Hz) based on the spectral observation. Then an SVM classifier, trained for the detection of a grid with a particular nominal frequency, narrows down the list of possible grids on the basis of different discriminating features extracted from the electric network frequency (ENF) signal. The decision of the SVM classifier is then passed to the PM classifier that detects the final grid based on the minimum distance between the estimated poles of test and training grids. Thus, we start from the problem of classifying grids with different nominal frequencies and simplify the problem of classification in three stages based on nominal frequency, SVM and finally using PM classifier. This cascaded system of classification ensures better accuracy (15.57% higher) compared to traditional ENF-based SVM classifiers described in the literature.

arXiv:1912.00195

SGAS: Sequential Greedy Architecture Search

Architecture design has become a crucial component of successful deep learning. Recent progress in automatic neural architecture search (NAS) shows a lot of promise. However, discovered architectures often fail to generalize in the final evaluation. Architectures with a higher validation accuracy during the search phase may perform worse in the evaluation. Aiming to alleviate this common issue, we introduce sequential greedy architecture search (SGAS), an efficient method for neural architecture search. By dividing the search procedure into sub-problems, SGAS chooses and prunes candidate operations in a greedy fashion. We apply SGAS to search architectures for Convolutional Neural Networks (CNN) and Graph Convolutional Networks (GCN). Extensive experiments show that SGAS is able to find state-of-the-art architectures for tasks such as image classification, point cloud classification and node classification in protein-protein interaction graphs with minimal computational cost. Please visit https://www.deepgcns.org/auto/sgas for more information about SGAS.

arXiv:1912.00835

Low Rank Factorization for Compact Multi-Head Self-Attention

Effective representation learning from text has been an active area of research in the fields of NLP and text mining. Attention mechanisms have been at the forefront in order to learn contextual sentence representations. Current state-of-the-art approaches for many NLP tasks use large pre-trained language models such as BERT, XLNet and so on for learning representations. These models are based on the Transformer architecture that involves recurrent blocks of computation consisting of multi-head self-attention and feedforward networks. One of the major bottlenecks largely contributing to the computational complexity of the Transformer models is the self-attention layer, that is both computationally expensive and parameter intensive. In this work, we introduce a novel multi-head self-attention mechanism operating on GRUs that is shown to be computationally cheaper and more parameter efficient than self-attention mechanism proposed in Transformers for text classification tasks. The efficiency of our approach mainly stems from two optimizations; 1) we use low-rank matrix factorization of the affinity matrix to efficiently get multiple attention distributions instead of having separate parameters for each head 2) attention scores are obtained by querying a global context vector instead of densely querying all the words in the sentence. We evaluate the performance of the proposed model on tasks such as sentiment analysis from movie reviews, predicting business ratings from reviews and classifying news articles into topics. We find that the proposed approach matches or outperforms a series of strong baselines and is more parameter efficient than comparable multi-head approaches. We also perform qualitative analyses to verify that the proposed approach is interpretable and captures context-dependent word importance.

arXiv:1912.00819공식 코드 · Papers with Code

EDA: Enriching Emotional Dialogue Acts using an Ensemble of Neural Annotators

The recognition of emotion and dialogue acts enriches conversational analysis and help to build natural dialogue systems. Emotion interpretation makes us understand feelings and dialogue acts reflect the intentions and performative functions in the utterances. However, most of the textual and multi-modal conversational emotion corpora contain only emotion labels but not dialogue acts. To address this problem, we propose to use a pool of various recurrent neural models trained on a dialogue act corpus, with and without context. These neural models annotate the emotion corpora with dialogue act labels, and an ensemble annotator extracts the final dialogue act label. We annotated two accessible multi-modal emotion corpora: IEMOCAP and MELD. We analyzed the co-occurrence of emotion and dialogue act labels and discovered specific relations. For example, Accept/Agree dialogue acts often occur with the Joy emotion, Apology with Sadness, and Thanking with Joy. We make the Emotional Dialogue Acts (EDA) corpus publicly available to the research community for further study and analysis.

arXiv:1912.00544

Multi-Scale Self-Attention for Text Classification

In this paper, we introduce the prior knowledge, multi-scale structure, into self-attention modules. We propose a Multi-Scale Transformer which uses multi-scale multi-head self-attention to capture features from different scales. Based on the linguistic perspective and the analysis of pre-trained Transformer (BERT) on a huge corpus, we further design a strategy to control the scale distribution for each layer. Results of three different kinds of tasks (21 datasets) show our Multi-Scale Transformer outperforms the standard Transformer consistently and significantly on small and moderate size datasets.

arXiv:1912.00509공식 코드 · Papers with Code

Speeding up Word Mover's Distance and its variants via properties of distances between embeddings

The Word Mover's Distance (WMD) proposed by Kusner et al. is a distance between documents that takes advantage of semantic relations among words that are captured by their embeddings. This distance proved to be quite effective, obtaining state-of-art error rates for classification tasks, but is also impracticable for large collections/documents due to its computational complexity. For circumventing this problem, variants of WMD have been proposed. Among them, Relaxed Word Mover's Distance (RWMD) is one of the most successful due to its simplicity, effectiveness, and also because of its fast implementations. Relying on assumptions that are supported by empirical properties of the distances between embeddings, we propose an approach to speed up both WMD and RWMD. Experiments over 10 datasets suggest that our approach leads to a significant speed-up in document classification tasks while maintaining the same error rates.

arXiv:1912.00127

A Hybrid Approach Towards Two Stage Bengali Question Classification Utilizing Smart Data Balancing Technique

Question classification (QC) is the primary step of the Question Answering (QA) system. Question Classification (QC) system classifies the questions in particular classes so that Question Answering (QA) System can provide correct answers for the questions. Our system categorizes the factoid type questions asked in natural language after extracting features of the questions. We present a two stage QC system for Bengali. It utilizes one dimensional convolutional neural network for classifying questions into coarse classes in the first stage. Word2vec representation of existing words of the question corpus have been constructed and used for assisting 1D CNN. A smart data balancing technique has been employed for giving data hungry convolutional neural network the advantage of a greater number of effective samples to learn from. For each coarse class, a separate Stochastic Gradient Descent (SGD) based classifier has been used in order to differentiate among the finer classes within that coarse class. TF-IDF representation of each word has been used as feature for the SGD classifiers implemented as part of second stage classification. Experiments show the effectiveness of our proposed method for Bengali question classification.

arXiv:1912.00113

Tag Recommendation by Word-Level Tag Sequence Modeling

In this paper, we transform tag recommendation into a word-based text generation problem and introduce a sequence-to-sequence model. The model inherits the advantages of LSTM-based encoder for sequential modeling and attention-based decoder with local positional encodings for learning relations globally. Experimental results on Zhihu datasets illustrate the proposed model outperforms other state-of-the-art text classification based methods.

arXiv:1911.13257

Non-Intrusive Electrical Appliances Monitoring and Classification using K-Nearest Neighbors

Non-Intrusive Load Monitoring (NILM) is the method of detecting an individual device's energy signal from an aggregated energy consumption signature [1]. As existing energy meters provide very little to no information regarding the energy consumption of individual appliances apart from the aggregated power rating, the spotting of individual appliances' energy usages by NILM will not only provide consumers the feedback of appliance-specific energy usage but also lead to the changes of their consumption behavior which facilitate energy conservation. B Neenan et al. [2] have demonstrated that direct individual appliance-specific energy usage signals lead to consumers' behavioral changes which improves energy efficiency by as much as 15%. Upon disaggregation of an energy signal, the signal needs to be classified according to the appropriate appliance. Hence, the goal of this paper is to disaggregate total energy consumption data to individual appliance signature and then classify appliance-specific energy loads using a prominent supervised classification method known as K-Nearest Neighbors (KNN). To perform this operation we have used a publicly accessible dataset of power signals from several houses known as the REDD dataset. Before applying KNN, data is preprocessed for each device. Then KNN is applied to check whether their energy consumption signature is separable or not. KNN is applied with K=5.

arXiv:1911.12478공식 코드 · Papers with Code

Metre as a stylometric feature in Latin hexameter poetry

This paper demonstrates that metre is a privileged indicator of authorial style in classical Latin hexameter poetry. Using only metrical features, pairwise classification experiments are performed between 5 first-century authors (10 comparisons) using four different machine-learning models. The results showed a two-label classification accuracy of at least 95% with samples as small as ten lines and no greater than eighty lines (up to around 500 words). These sample sizes are an order of magnitude smaller than those typically recommended for BOW ('bag of words') or n-gram approaches, and the reported accuracy is outstanding. Additionally, this paper explores the potential for novelty (forgery) detection, or 'one-class classification'. An analysis of the disputed Aldine Additamentum (Sil. Ital. Puni. 8:144-225) concludes (p=0.0013) that the metrical style differs significantly from that of the rest of the poem.

arXiv:1911.12446

QubitHD: A Stochastic Acceleration Method for HD Computing-Based Machine Learning

Machine Learning algorithms based on Brain-inspired Hyperdimensional(HD) computing imitate cognition by exploiting statistical properties of high-dimensional vector spaces. It is a promising solution for achieving high energy efficiency in different machine learning tasks, such as classification, semi-supervised learning, and clustering. A weakness of existing HD computing-based ML algorithms is the fact that they have to be binarized to achieve very high energy efficiency. At the same time, binarized models reach lower classification accuracies. To solve the problem of the trade-off between energy efficiency and classification accuracy, we propose the QubitHD algorithm. It stochastically binarizes HD-based algorithms, while maintaining comparable classification accuracies to their non-binarized counterparts. The FPGA implementation of QubitHD provides a 65% improvement in terms of energy efficiency, and a 95% improvement in terms of training time, as compared to state-of-the-art HD-based ML algorithms. It also outperforms state-of-the-art low-cost classifiers (such as Binarized Neural Networks) in terms of speed and energy efficiency by an order of magnitude during training and inference.

arXiv:1911.12425

Learning with less data via Weakly Labeled Patch Classification in Digital Pathology

In Digital Pathology (DP), labeled data is generally very scarce due to the requirement that medical experts provide annotations. We address this issue by learning transferable features from weakly labeled data, which are collected from various parts of the body and are organized by non-medical experts. In this paper, we show that features learned from such weakly labeled datasets are indeed transferable and allow us to achieve highly competitive patch classification results on the colorectal cancer (CRC) dataset [1] and the PatchCamelyon (PCam) dataset [2] while using an order of magnitude less labeled data.

arXiv:1911.12903

Land Cover Change Detection via Semantic Segmentation

This paper presents a change detection method that identifies land cover changes from aerial imagery, using semantic segmentation, a machine learning approach. We present a land cover classification training pipeline with Deeplab v3+, state-of-the-art semantic segmentation technology, including data preparation, model training for seven land cover types, and model exporting modules. In the land cover change detection system, the inputs are images retrieved from Google Earth at the same location but from different times. The system then predicts semantic segmentation results on these images using the trained model and calculates the land cover class percentage for each input image. We see an improvement in the accuracy of the land cover semantic segmentation model, with a mean IoU of 0.756 compared to 0.433, as reported in the DeepGlobe land cover classification challenge. The land cover change detection system that leverages the state-of-the-art semantic segmentation technology is proposed and can be used for deforestation analysis, land management, and urban planning.

arXiv:1911.12885공식 코드 · Papers with Code

Geometric Back-projection Network for Point Cloud Classification

As the basic task of point cloud analysis, classification is fundamental but always challenging. To address some unsolved problems of existing methods, we propose a network that captures geometric features of point clouds for better representations. To achieve this, on the one hand, we enrich the geometric information of points in low-level 3D space explicitly. On the other hand, we apply CNN-based structures in high-level feature spaces to learn local geometric context implicitly. Specifically, we leverage an idea of error-correcting feedback structure to capture the local features of point clouds comprehensively. Furthermore, an attention module based on channel affinity assists the feature map to avoid possible redundancy by emphasizing its distinct channels. The performance on both synthetic and real-world point clouds datasets demonstrate the superiority and applicability of our network. Comparing with other state-of-the-art methods, our approach balances accuracy and efficiency.

arXiv:1911.12688

A novel classification-selection approach for the self updating of template-based face recognition systems

The boosting on the need of security notably increased the amount of possible facial recognition applications, especially due to the success of the Internet of Things (IoT) paradigm. However, although handcrafted and deep learning-inspired facial features reached a significant level of compactness and expressive power, the facial recognition performance still suffers from intra-class variations such as ageing, facial expressions, lighting changes, and pose. These variations cannot be captured in a single acquisition and require multiple acquisitions of long duration, which are expensive and need a high level of collaboration from the users. Among others, self-update algorithms have been proposed in order to mitigate these problems. Self-updating aims to add novel templates to the users' gallery among the inputs submitted during system operations. Consequently, computational complexity and storage space tend to be among the critical requirements of these algorithms. The present paper deals with the above problems by a novel template-based self-update algorithm, able to keep over time the expressive power of a limited set of templates stored in the system database. The rationale behind the proposed approach is in the working hypothesis that a dominating mode characterises the features' distribution given the client. Therefore, the key point is to select the best templates around that mode. We propose two methods, which are tested on systems based on handcrafted features and deep-learning-inspired autoencoders at the state-of-the-art. Three benchmark data sets are used. Experimental results confirm that, by effective and compact feature sets which can support our working hypothesis, the proposed classification-selection approaches overcome the problem of manual updating and, in case, stringent computational requirements.