arXiv subjects

Yifan Hu

Publications and source records attributed to Yifan Hu.

At least 127 records · Page 7Linked to original sources

arXiv2021

DeepGD: A Deep Learning Framework for Graph Drawing Using GNN

In the past decades, many graph drawing techniques have been proposed for generating aesthetically pleasing graph layouts. However, it remains a challenging task since different layout methods tend to highlight different characteristics of the graphs. Recently, studies on deep learning based graph drawing algorithm have emerged but they are often not generalizable to arbitrary graphs without re-training. In this paper, we propose a Convolutional Graph Neural Network based deep learning framework, DeepGD, which can draw arbitrary graphs once trained. It attempts to generate layouts by compromising among multiple pre-specified aesthetics considering a good graph layout usually complies with multiple aesthetics simultaneously. In order to balance the trade-off, we propose two adaptive training strategies which adjust the weight factor of each aesthetic dynamically during training. The quantitative and qualitative assessment of DeepGD demonstrates that it is capable of drawing arbitrary graphs effectively, while being flexible at accommodating different aesthetic criteria.

cs.LG↗

arXiv2021

Efficient Ring-topology Decentralized Federated Learning with Deep Generative Models for Industrial Artificial Intelligent

By leveraging deep learning based technologies, the data-driven based approaches have reached great success with the rapid increase of data generated of Industrial Indernet of Things(IIot). However, security and privacy concerns are obstacles for data providers in many sensitive data-driven industrial scenarios, such as healthcare and auto-driving. Many Federated Learning(FL) approaches have been proposed with DNNs for IIoT applications, these works still suffer from low usability of data due to data incompleteness, low quality, insufficient quantity, sensitivity, etc. Therefore, we propose a ring-topogy based decentralized federated learning(RDFL) scheme for Deep Generative Models(DGMs), where DGMs is a promising solution for solving the aforementioned data usability issues. Compare with existing IIoT FL works, our RDFL schemes provides communication efficiency and maintain training performance to boost DGMs in target IIoT tasks. A novel ring FL topology as well as a map-reduce based synchronizing method are designed in the proposed RDFL to improve decentralized FL performance and bandwidth utilization. In addition, InterPlanetary File System(IPFS) is introduced to further improve communication efficiency and FL security. Extensive experiments have been taken to demonstate the superiority of RDFL with either independent and identically distributed(IID) datasets or non-independent and identically distributed(Non-IID) datasets.

cs.LG↗

arXiv2021

GFL: A Decentralized Federated Learning Framework Based On Blockchain

Federated learning(FL) is a rapidly growing field and many centralized and decentralized FL frameworks have been proposed. However, it is of great challenge for current FL frameworks to improve communication performance and maintain the security and robustness under malicious node attacks. In this paper, we propose Galaxy Federated Learning Framework(GFL), a decentralized FL framework based on blockchain. GFL introduces the consistent hashing algorithm to improve communication performance and proposes a novel ring decentralized FL algorithm(RDFL) to improve decentralized FL performance and bandwidth utilization. In addition, GFL introduces InterPlanetary File System(IPFS) and blockchain to further improve communication efficiency and FL security. Our experiments show that GFL improves communication performance and decentralized FL performance under the data poisoning of malicious nodes and non-independent and identically distributed(Non-IID) datasets.

cs.LG↗

arXiv2021

What's in a Name? -- Gender Classification of Names with Character Based Machine Learning Models

Gender information is no longer a mandatory input when registering for an account at many leading Internet companies. However, prediction of demographic information such as gender and age remains an important task, especially in intervention of unintentional gender/age bias in recommender systems. Therefore it is necessary to infer the gender of those users who did not to provide this information during registration. We consider the problem of predicting the gender of registered users based on their declared name. By analyzing the first names of 100M+ users, we found that genders can be very effectively classified using the composition of the name strings. We propose a number of character based machine learning models, and demonstrate that our models are able to infer the gender of users with much higher accuracy than baseline models. Moreover, we show that using the last names in addition to the first names improves classification performance further.

cs.LG↗

arXiv2020

Political Posters Identification with Appearance-Text Fusion

In this paper, we propose a method that efficiently utilizes appearance features and text vectors to accurately classify political posters from other similar political images. The majority of this work focuses on political posters that are designed to serve as a promotion of a certain political event, and the automated identification of which can lead to the generation of detailed statistics and meets the judgment needs in a variety of areas. Starting with a comprehensive keyword list for politicians and political events, we curate for the first time an effective and practical political poster dataset containing 13K human-labeled political images, including 3K political posters that explicitly support a movement or a campaign. Second, we make a thorough case study for this dataset and analyze common patterns and outliers of political posters. Finally, we propose a model that combines the power of both appearance and text information to classify political posters with significantly high accuracy.

cs.CV↗

arXiv2020

Going Deeper With Directly-Trained Larger Spiking Neural Networks

Spiking neural networks (SNNs) are promising in a bio-plausible coding for spatio-temporal information and event-driven signal processing, which is very suited for energy-efficient implementation in neuromorphic hardware. However, the unique working mode of SNNs makes them more difficult to train than traditional networks. Currently, there are two main routes to explore the training of deep SNNs with high performance. The first is to convert a pre-trained ANN model to its SNN version, which usually requires a long coding window for convergence and cannot exploit the spatio-temporal features during training for solving temporal tasks. The other is to directly train SNNs in the spatio-temporal domain. But due to the binary spike activity of the firing function and the problem of gradient vanishing or explosion, current methods are restricted to shallow architectures and thereby difficult in harnessing large-scale datasets (e.g. ImageNet). To this end, we propose a threshold-dependent batch normalization (tdBN) method based on the emerging spatio-temporal backpropagation, termed "STBP-tdBN", enabling direct training of a very deep SNN and the efficient implementation of its inference on neuromorphic hardware. With the proposed method and elaborated shortcut connection, we significantly extend directly-trained SNNs from a shallow structure ( < 10 layer) to a very deep structure (50 layers). Furthermore, we theoretically analyze the effectiveness of our method based on "Block Dynamical Isometry" theory. Finally, we report superior accuracy results including 93.15 % on CIFAR-10, 67.8 % on DVS-CIFAR10, and 67.05% on ImageNet with very few timesteps. To our best knowledge, it's the first time to explore the directly-trained deep SNNs with high performance on ImageNet.

cs.NE↗

arXiv2020

Repulsive Attention: Rethinking Multi-head Attention as Bayesian Inference

The neural attention mechanism plays an important role in many natural language processing applications. In particular, the use of multi-head attention extends single-head attention by allowing a model to jointly attend information from different perspectives. Without explicit constraining, however, multi-head attention may suffer from attention collapse, an issue that makes different heads extract similar attentive features, thus limiting the model's representation power. In this paper, for the first time, we provide a novel understanding of multi-head attention from a Bayesian perspective. Based on the recently developed particle-optimization sampling techniques, we propose a non-parametric approach that explicitly improves the repulsiveness in multi-head attention and consequently strengthens model's expressiveness. Remarkably, our Bayesian interpretation provides theoretical inspirations on the not-well-understood questions: why and how one uses multi-head attention. Extensive experiments on various attention models and applications demonstrate that the proposed repulsive attention can improve the learned feature diversity, leading to more informative representations with consistent performance improvement on various tasks.

cs.LG↗

arXiv2020

HABERTOR: An Efficient and Effective Deep Hatespeech Detector

We present our HABERTOR model for detecting hatespeech in large scale user-generated content. Inspired by the recent success of the BERT model, we propose several modifications to BERT to enhance the performance on the downstream hatespeech classification task. HABERTOR inherits BERT's architecture, but is different in four aspects: (i) it generates its own vocabularies and is pre-trained from the scratch using the largest scale hatespeech dataset; (ii) it consists of Quaternion-based factorized components, resulting in a much smaller number of parameters, faster training and inferencing, as well as less memory usage; (iii) it uses our proposed multi-source ensemble heads with a pooling layer for separate input sources, to further enhance its effectiveness; and (iv) it uses a regularized adversarial training with our proposed fine-grained and adaptive noise magnitude to enhance its robustness. Through experiments on the large-scale real-world hatespeech dataset with 1.4M annotated comments, we show that HABERTOR works better than 15 state-of-the-art hatespeech detection methods, including fine-tuning Language Models. In particular, comparing with BERT, our HABERTOR is 4~5 times faster in the training/inferencing phase, uses less than 1/3 of the memory, and has better performance, even though we pre-train it by using less than 1% of the number of words. Our generalizability analysis shows that HABERTOR transfers well to other unseen hatespeech datasets and is a more efficient and effective alternative to BERT for the hatespeech classification.

cs.CL↗

arXiv2020

Comprehensive SNN Compression Using ADMM Optimization and Activity Regularization

As well known, the huge memory and compute costs of both artificial neural networks (ANNs) and spiking neural networks (SNNs) greatly hinder their deployment on edge devices with high efficiency. Model compression has been proposed as a promising technique to improve the running efficiency via parameter and operation reduction. Whereas, this technique is mainly practiced in ANNs rather than SNNs. It is interesting to answer how much an SNN model can be compressed without compromising its functionality, where two challenges should be addressed: i) the accuracy of SNNs is usually sensitive to model compression, which requires an accurate compression methodology; ii) the computation of SNNs is event-driven rather than static, which produces an extra compression dimension on dynamic spikes. To this end, we realize a comprehensive SNN compression through three steps. First, we formulate the connection pruning and weight quantization as a constrained optimization problem. Second, we combine spatio-temporal backpropagation (STBP) and alternating direction method of multipliers (ADMM) to solve the problem with minimum accuracy loss. Third, we further propose activity regularization to reduce the spike events for fewer active operations. These methods can be applied in either a single way for moderate compression or a joint way for aggressive compression. We define several quantitative metrics to evaluation the compression performance for SNNs. Our methodology is validated in pattern recognition tasks over MNIST, N-MNIST, CIFAR10, and CIFAR100 datasets, where extensive comparisons, analyses, and insights are provided. To our best knowledge, this is the first work that studies SNN compression in a comprehensive manner by exploiting all compressible components and achieves better results.

cs.NE↗

arXiv2020

Comparing to Learn: Surpassing ImageNet Pretraining on Radiographs By Comparing Image Representations

In deep learning era, pretrained models play an important role in medical image analysis, in which ImageNet pretraining has been widely adopted as the best way. However, it is undeniable that there exists an obvious domain gap between natural images and medical images. To bridge this gap, we propose a new pretraining method which learns from 700k radiographs given no manual annotations. We call our method as Comparing to Learn (C2L) because it learns robust features by comparing different image representations. To verify the effectiveness of C2L, we conduct comprehensive ablation studies and evaluate it on different tasks and datasets. The experimental results on radiographs show that C2L can outperform ImageNet pretraining and previous state-of-the-art approaches significantly. Code and models are available.

cs.CV↗

arXiv2020

Embedding Task Knowledge into 3D Neural Networks via Self-supervised Learning

Deep learning highly relies on the amount of annotated data. However, annotating medical images is extremely laborious and expensive. To this end, self-supervised learning (SSL), as a potential solution for deficient annotated data, attracts increasing attentions from the community. However, SSL approaches often design a proxy task that is not necessarily related to target task. In this paper, we propose a novel SSL approach for 3D medical image classification, namely Task-related Contrastive Prediction Coding (TCPC), which embeds task knowledge into training 3D neural networks. The proposed TCPC first locates the initial candidate lesions via supervoxel estimation using simple linear iterative clustering. Then, we extract features from the sub-volume cropped around potential lesion areas, and construct a calibrated contrastive predictive coding scheme for self-supervised learning. Extensive experiments are conducted on public and private datasets. The experimental results demonstrate the effectiveness of embedding lesion-related prior-knowledge into neural networks for 3D medical image classification.

cs.CV↗

arXiv2020

Multi-Modality Generative Adversarial Networks with Tumor Consistency Loss for Brain MR Image Synthesis

Magnetic Resonance (MR) images of different modalities can provide complementary information for clinical diagnosis, but whole modalities are often costly to access. Most existing methods only focus on synthesizing missing images between two modalities, which limits their robustness and efficiency when multiple modalities are missing. To address this problem, we propose a multi-modality generative adversarial network (MGAN) to synthesize three high-quality MR modalities (FLAIR, T1 and T1ce) from one MR modality T2 simultaneously. The experimental results show that the quality of the synthesized images by our proposed methods is better than the one synthesized by the baseline model, pix2pix. Besides, for MR brain image synthesis, it is important to preserve the critical tumor information in the generated modalities, so we further introduce a multi-modality tumor consistency loss to MGAN, called TC-MGAN. We use the synthesized modalities by TC-MGAN to boost the tumor segmentation accuracy, and the results demonstrate its effectiveness.

cs.CV↗

arXiv2020

Sample Complexity of Sample Average Approximation for Conditional Stochastic Optimization

In this paper, we study a class of stochastic optimization problems, referred to as the \emph{Conditional Stochastic Optimization} (CSO), in the form of $\min_{x \in \mathcal{X}} \EE_ξf_ξ\Big({\EE_{η|ξ}[g_η(x,ξ)]}\Big)$, which finds a wide spectrum of applications including portfolio selection, reinforcement learning, robust learning, causal inference and so on. Assuming availability of samples from the distribution $\PP(ξ)$ and samples from the conditional distribution $\PP(η|ξ)$, we establish the sample complexity of the sample average approximation (SAA) for CSO, under a variety of structural assumptions, such as Lipschitz continuity, smoothness, and error bound conditions. We show that the total sample complexity improves from $\cO(d/\eps^4)$ to $\cO(d/\eps^3)$ when assuming smoothness of the outer function, and further to $\cO(1/\eps^2)$ when the empirical function satisfies the quadratic growth condition. We also establish the sample complexity of a modified SAA, when $ξ$ and $η$ are independent. Several numerical experiments further support our theoretical findings. Keywords: stochastic optimization, sample average approximation, large deviations theory

math.OC↗

arXiv2020

A Deep Structural Model for Analyzing Correlated Multivariate Time Series

Multivariate time series are routinely encountered in real-world applications, and in many cases, these time series are strongly correlated. In this paper, we present a deep learning structural time series model which can (i) handle correlated multivariate time series input, and (ii) forecast the targeted temporal sequence by explicitly learning/extracting the trend, seasonality, and event components. The trend is learned via a 1D and 2D temporal CNN and LSTM hierarchical neural net. The CNN-LSTM architecture can (i) seamlessly leverage the dependency among multiple correlated time series in a natural way, (ii) extract the weighted differencing feature for better trend learning, and (iii) memorize the long-term sequential pattern. The seasonality component is approximated via a non-liner function of a set of Fourier terms, and the event components are learned by a simple linear function of regressor encoding the event dates. We compare our model with several state-of-the-art methods through a comprehensive set of experiments on a variety of time series data sets, such as forecasts of Amazon AWS Simple Storage Service (S3) and Elastic Compute Cloud (EC2) billings, and the closing prices for corporate stocks in the same category.

stat.ML↗

arXiv2020

Large-scale Gender/Age Prediction of Tumblr Users

Tumblr, as a leading content provider and social media, attracts 371 million monthly visits, 280 million blogs and 53.3 million daily posts. The popularity of Tumblr provides great opportunities for advertisers to promote their products through sponsored posts. However, it is a challenging task to target specific demographic groups for ads, since Tumblr does not require user information like gender and ages during their registration. Hence, to promote ad targeting, it is essential to predict user's demography using rich content such as posts, images and social connections. In this paper, we propose graph based and deep learning models for age and gender predictions, which take into account user activities and content features. For graph based models, we come up with two approaches, network embedding and label propagation, to generate connection features as well as directly infer user's demography. For deep learning models, we leverage convolutional neural network (CNN) and multilayer perceptron (MLP) to prediction users' age and gender. Experimental results on real Tumblr daily dataset, with hundreds of millions of active users and billions of following relations, demonstrate that our approaches significantly outperform the baseline model, by improving the accuracy relatively by 81% for age, and the AUC and accuracy by 5\% for gender.

cs.LG↗

arXiv2019

Self-supervised Feature Learning for 3D Medical Images by Playing a Rubik's Cube

Witnessed the development of deep learning, increasing number of studies try to build computer aided diagnosis systems for 3D volumetric medical data. However, as the annotations of 3D medical data are difficult to acquire, the number of annotated 3D medical images is often not enough to well train the deep learning networks. The self-supervised learning deeply exploiting the information of raw data is one of the potential solutions to loose the requirement of training data. In this paper, we propose a self-supervised learning framework for the volumetric medical images. A novel proxy task, i.e., Rubik's cube recovery, is formulated to pre-train 3D neural networks. The proxy task involves two operations, i.e., cube rearrangement and cube rotation, which enforce networks to learn translational and rotational invariant features from raw 3D data. Compared to the train-from-scratch strategy, fine-tuning from the pre-trained network leads to a better accuracy on various tasks, e.g., brain hemorrhage classification and brain tumor segmentation. We show that our self-supervised learning approach can substantially boost the accuracies of 3D deep learning networks on the volumetric medical datasets without using extra data. To our best knowledge, this is the first work focusing on the self-supervised learning of 3D neural networks.

cs.CV↗

arXiv2019

A GLCM Embedded CNN Strategy for Computer-aided Diagnosis in Intracerebral Hemorrhage

Computer-aided diagnosis (CADx) systems have been shown to assist radiologists by providing classifications of all kinds of medical images like Computed tomography (CT) and Magnetic resonance (MR). Currently, convolutional neural networks play an important role in CADx. However, since CNN model should have a square-like input, it is usually difficult to directly apply the CNN algorithms on the irregular segmentation region of interests (ROIs) where the radiologists are interested in. In this paper, we propose a new approach to construct the model by extracting and converting the information of the irregular region into a fixed-size Gray-Level Co-Occurrence Matrix (GLCM) and then utilize the GLCM as one input of our CNN model. In this way, as an useful implementary to the original CNN, a couple of GLCM-based features are also extracted by CNN. Meanwhile, the network will pay more attention to the important lesion area and achieve a higher accuracy in classification. Experiments are performed on three classification databases: Hemorrhage, BraTS18 and Cervix to validate the universality of our innovative model. In conclusion, the proposed framework outperforms the corresponding state-of-art algorithms on each database with both test losses and classification accuracy as the evaluation criteria.

cs.CV↗

arXiv2017

HARP: Hierarchical Representation Learning for Networks

We present HARP, a novel method for learning low dimensional embeddings of a graph's nodes which preserves higher-order structural features. Our proposed method achieves this by compressing the input graph prior to embedding it, effectively avoiding troublesome embedding configurations (i.e. local minima) which can pose problems to non-convex optimization. HARP works by finding a smaller graph which approximates the global structure of its input. This simplified graph is used to learn a set of initial representations, which serve as good initializations for learning representations in the original, detailed graph. We inductively extend this idea, by decomposing a graph in a series of levels, and then embed the hierarchy of graphs from the coarsest one to the original graph. HARP is a general meta-strategy to improve all of the state-of-the-art neural algorithms for embedding graphs, including DeepWalk, LINE, and Node2vec. Indeed, we demonstrate that applying HARP's hierarchical paradigm yields improved implementations for all three of these methods, as evaluated on both classification tasks on real-world graphs such as DBLP, BlogCatalog, CiteSeer, and Arxiv, where we achieve a performance gain over the original implementations by up to 14% Macro F1.

cs.SI↗