Xiang Zhang

Bio: Xiang Zhang is an academic researcher from Baylor College of Medicine. The author has contributed to research in topics: Medicine & Computer science. The author has an hindex of 154, co-authored 1733 publications receiving 117576 citations. Previous affiliations of Xiang Zhang include University of California, Berkeley & University of Texas MD Anderson Cancer Center.

Deep Neural Network Hyperparameter Optimization with Orthogonal Array Tuning

Abstract: Deep learning algorithms have achieved excellent performance lately in a wide range of fields (e.g., computer version). However, a severe challenge faced by deep learning is the high dependency on hyper-parameters. The algorithm results may fluctuate dramatically under the different configuration of hyper-parameters. Addressing the above issue, this paper presents an efficient Orthogonal Array Tuning Method (OATM) for deep learning hyper-parameter tuning. We describe the OATM approach in five detailed steps and elaborate on it using two widely used deep neural network structures (Recurrent Neural Networks and Convolutional Neural Networks). The proposed method is compared to the state-of-the-art hyper-parameter tuning methods including manually (e.g., grid search and random search) and automatically (e.g., Bayesian Optimization) ones. The experiment results state that OATM can significantly save the tuning time compared to the state-of-the-art methods while preserving the satisfying performance.

Diffusion-limited photopolymerization in scanning micro-stereolithography

Abstract: The trade-off between process speed and resolution in microstereolithography (μSL) roots on the diffusion-limited kinetics of photopolymerization. Using a numerical model, we have investigated the influence of diffusion dominant effect under high photon flux. Radical depletion turned out to limit the smallest feature achievable to the order of 10 μm under high process speed. A solution of pulsed laser curing is proposed in order to realize sub-micron resolution in high speed μSL process.

miR-141-Mediated Regulation of Brain Metastasis From Breast Cancer

Abstract: BACKGROUND Brain metastasis poses a major treatment challenge and remains an unmet clinical need. Finding novel therapies to prevent and treat brain metastases requires an understanding of the biology and molecular basis of the process, which currently is constrained by a dearth of experimental models and specific therapeutic targets. METHODS Green Fluorescent Protein (GFP)-labeled breast cancer cells were injected via tail vein into SCID/Beige mice (n = 10-15 per group), and metastatic colonization to the brain and lung was evaluated eight weeks later. Knockdown and overexpression of miR-141 were achieved with lentiviral vectors. Serum levels of miR-141 were measured from breast cancer patients (n = 105), and the association with clinical outcome was determined by Kaplan-Meier method. All statistical tests were two-sided. RESULTS Novel brain metastasis mouse models were developed via tail vein injection of parental triple-negative and human epidermal growth factor receptor 2 (HER2)-overexpressing inflammatory breast cancer lines. Knockdown of miR-141 inhibited metastatic colonization to brain (miR-141 knockdown vs control: SUM149, 0/8 mice vs 6/9 mice,P= .009; MDA-IBC3, 2/14 mice vs 10/15 mice,P= .007). Ectopic expression of miR-141 in nonexpressing MDA-MB-231 enhanced brain metastatic colonization (5/9 mice vs 0/10 mice,P= .02). Furthermore, high miR-141 serum levels were associated with shorter brain metastasis-free survival (P= .04) and were an independent predictor of progression-free survival (hazard ratio [HR] = 4.77, 95% confidence interval [CI] = 2.61 to 8.71,P< .001) and overall survival (HR = 7.22, 95% CI = 3.46 to 15.06,P< .001). CONCLUSIONS Our study suggests miR-141 is a regulator of brain metastasis from breast cancer and should be examined as a biomarker and potential target to prevent and treat brain metastases.

EEG-based motion intention recognition via multi-task RNNs

Abstract: Recognition of human intention based on Electroen-cephalography (EEG) signals attracts strong research interest in pattern recognition because of its promising applications that enable non-muscular communications and controls. Over the past few years, most EEG-based recognition works make significant efforts to learn ex-tracted features to explore specific patterns between a segment of EEG signals and the corresponding activi-ties. Unfortunately, vectorization-based feature repre-sentations, either vector-like or matrix-like ones, suffer from massive signal noise and difficulties of exploiting signal correlations between adjacent sensors of EEG sig-nals. Most importantly, EEG signals are represented by one unique frequency and then fed into the subse-quent learning model. Neglecting different frequencies of EEG signals can be detrimental to activity recogni-tion because a particular frequency of EEG signals is more helpful to recognize some activities. Inspired by this idea, we propose to extract EEG signals with different frequencies and introduce a novel Multi-task deep learning model to learn the human intentions. We have conducted extensive experiments on a publicly avail-able EEG benchmark dataset and compared our method with many state-of-the-art algorithms. The experimen-tal results demonstrate that the proposed Multi-task deep recurrent neural network outperforms all the com-pared methods in a multi-class scenario.

Metric Learning from Relative Comparisons by Minimizing Squared Residual

Abstract: Recent studies [1] -- [5] have suggested using constraints in the form of relative distance comparisons to represent domain knowledge: d(a, b)

Deep Residual Learning for Image Recognition

Abstract: Deeper neural networks are more difficult to train. We present a residual learning framework to ease the training of networks that are substantially deeper than those used previously. We explicitly reformulate the layers as learning residual functions with reference to the layer inputs, instead of learning unreferenced functions. We provide comprehensive empirical evidence showing that these residual networks are easier to optimize, and can gain accuracy from considerably increased depth. On the ImageNet dataset we evaluate residual nets with a depth of up to 152 layers—8× deeper than VGG nets [40] but still having lower complexity. An ensemble of these residual nets achieves 3.57% error on the ImageNet test set. This result won the 1st place on the ILSVRC 2015 classification task. We also present analysis on CIFAR-10 with 100 and 1000 layers. The depth of representations is of central importance for many visual recognition tasks. Solely due to our extremely deep representations, we obtain a 28% relative improvement on the COCO object detection dataset. Deep residual nets are foundations of our submissions to ILSVRC & COCO 2015 competitions1, where we also won the 1st places on the tasks of ImageNet detection, ImageNet localization, COCO detection, and COCO segmentation.

Very Deep Convolutional Networks for Large-Scale Image Recognition

Abstract: In this work we investigate the effect of the convolutional network depth on its accuracy in the large-scale image recognition setting. Our main contribution is a thorough evaluation of networks of increasing depth using an architecture with very small (3x3) convolution filters, which shows that a significant improvement on the prior-art configurations can be achieved by pushing the depth to 16-19 weight layers. These findings were the basis of our ImageNet Challenge 2014 submission, where our team secured the first and the second places in the localisation and classification tracks respectively. We also show that our representations generalise well to other datasets, where they achieve state-of-the-art results. We have made our two best-performing ConvNet models publicly available to facilitate further research on the use of deep visual representations in computer vision.

Deep Residual Learning for Image Recognition

Abstract: Deeper neural networks are more difficult to train. We present a residual learning framework to ease the training of networks that are substantially deeper than those used previously. We explicitly reformulate the layers as learning residual functions with reference to the layer inputs, instead of learning unreferenced functions. We provide comprehensive empirical evidence showing that these residual networks are easier to optimize, and can gain accuracy from considerably increased depth. On the ImageNet dataset we evaluate residual nets with a depth of up to 152 layers---8x deeper than VGG nets but still having lower complexity. An ensemble of these residual nets achieves 3.57% error on the ImageNet test set. This result won the 1st place on the ILSVRC 2015 classification task. We also present analysis on CIFAR-10 with 100 and 1000 layers. The depth of representations is of central importance for many visual recognition tasks. Solely due to our extremely deep representations, we obtain a 28% relative improvement on the COCO object detection dataset. Deep residual nets are foundations of our submissions to ILSVRC & COCO 2015 competitions, where we also won the 1st places on the tasks of ImageNet detection, ImageNet localization, COCO detection, and COCO segmentation.

Going deeper with convolutions

Abstract: We propose a deep convolutional neural network architecture codenamed Inception that achieves the new state of the art for classification and detection in the ImageNet Large-Scale Visual Recognition Challenge 2014 (ILSVRC14). The main hallmark of this architecture is the improved utilization of the computing resources inside the network. By a carefully crafted design, we increased the depth and width of the network while keeping the computational budget constant. To optimize quality, the architectural decisions were based on the Hebbian principle and the intuition of multi-scale processing. One particular incarnation used in our submission for ILSVRC14 is called GoogLeNet, a 22 layers deep network, the quality of which is assessed in the context of classification and detection.