15 research outputs found
Federated Analysis in COINSTAC Reveals Functional Network Connectivity and Spectral Links to Smoking and Alcohol Consumption in Nearly 2,000 Adolescent Brains
No full textNational Institutes of Health http://dx.doi.org/10.13039/100000002National Science Foundation http://dx.doi.org/10.13039/100000001Horizon 2020 Framework Programme http://dx.doi.org/10.13039/100010661Medical Research FoundationNational Institute for Health Research http://dx.doi.org/10.13039/501100000272National Institute for Health Research http://dx.doi.org/10.13039/501100000272National Institute for Health Research http://dx.doi.org/10.13039/501100000272Medical Research Council http://dx.doi.org/10.13039/501100000265Sixth Framework Programme http://dx.doi.org/10.13039/100011103Human Brain ProjectMedical Research Council http://dx.doi.org/10.13039/50110000026
Advances in Deep Learning through Gradient Amplification and Applications
No full textDeep neural networks currently play a prominent role in solving problems across a wide variety of disciplines. Improving performance of deep learning models and reducing their training times are some of the ongoing challenges. Increasing the depth of the networks improves performance but suffers from the problem of vanishing gradients and increased training times. In this research, we design methods to address these challenges in deep neural networks and demonstrate deep learning applications in several domains. We propose a gradient amplification based approach to train deep neural networks, which improves their training and testing accuraries, addresses vanishing gradients, as well as reduces the training time by reaching higher accuracies even at higher learning rates. We also develop an integrated training strategy to enable/disable amplification at certain epochs. Detailed analysis is performed on different neural networks using random amplification, where the layers to be amplified are selected randomly. The implications of gradient amplification on the number of layers, types of layers, amplification factors, training strategies and learning rates are studied in detail. With this knowledge, effective ways to update gradients are designed to perform amplification at layer-level and also at neuron-level. Lastly, we provide applications of deep learning methods to some of the challenging problems in the areas of smartgrids and bioinformatics. Deep neural networks with feed forward architectures are used to solve data integrity attacks in smart grids. We propose an image based preprocessing method to convert heterogenous genomic sequences into images which are then classified to detect Hepatitis C virus(HCV) infection stages. In summary, this research advances deep learning techniques and their applications to real world problems
Analysis of heterogeneous genomic samples using image normalization and machine learning
No full textBackground: Analysis of heterogeneous populations such as viral quasispecies is one of the most challenging bioinformatics problems. Although machine learning models are becoming to be widely employed for analysis of sequence data from such populations, their straightforward application is impeded by multiple challenges associated with technological limitations and biases, difficulty of selection of relevant features and need to compare genomic datasets of different sizes and structures. Results: We propose a novel preprocessing approach to transform irregular genomic data into normalized image data. Such representation allows to restate the problems of classification and comparison of heterogeneous populations as image classification problems which can be solved using variety of available machine learning tools. We then apply the proposed approach to two important problems in molecular epidemiology: inference of viral infection stage and detection of viral transmission clusters using next-generation sequencing data. The infection staging method has been applied to HCV HVR1 samples collected from 108 recently and 257 chronically infected individuals. The SVM-based image classification approach achieved more than 95% accuracy for both recently and chronically HCV-infected individuals. Clustering has been performed on the data collected from 33 epidemiologically curated outbreaks, yielding more than 97% accuracy. Conclusions: Sequence image normalization method allows for a robust conversion of genomic data into numerical data and overcomes several issues associated with employing machine learning methods to viral populations. Image data also help in the visualization of genomic data. Experimental results demonstrate that the proposed method can be successfully applied to different problems in molecular epidemiology and surveillance of viral diseases. Simple binary classifiers and clustering techniques applied to the image data are equally or more accurate than other models. © 2020, The Author(s)
Infant Sound Classification on Multi-stage CNNs with Hybrid Features and Prior Knowledge
No full textDeep Learning for Asphyxiated Infant Cry Classification Based on Acoustic Features and Weighted Prosodic Features
No full textA Survey on Algorithms for Intelligent Computing and Smart City Applications
Get PDFWith the rapid development of human society, the urbanization of the world’s population is also progressing rapidly. Urbanization has brought many challenges and problems to the development of cities. For example, the urban population is under excessive pressure, various natural resources and energy are increasingly scarce, and environmental pollution is increasing, etc. However, the original urban model has to be changed to enable people to live in greener and more sustainable cities, thus providing them with a more convenient and comfortable living environment. The new urban framework, the smart city, provides excellent opportunities to meet these challenges, while solving urban problems at the same time. At this stage, many countries are actively responding to calls for smart city development plans. This paper investigates the current stage of the smart city. First, it introduces the background of smart city development and gives a brief definition of the concept of the smart city. Second, it describes the framework of a smart city in accordance with the given definition. Finally, various intelligent algorithms to make cities smarter, along with specific examples, are discussed and analyzed
Multi‐view learning for benign epilepsy with centrotemporal spikes
No full textBenign epilepsy with centrotemporal spikes (BECT) may be the most popular epilepsy to attack children. In recent years, more and more studies have shown that magnetic resonance imaging (MRI) and functional magnetic resonance imaging (fMRI) are promising techniques in distinguishing BECT patients from healthy controls. However, these existing works have suffered from two limitations. On the one hand, they have paid more attention to the brain changes between BETC and healthy controls than developing machine learning methods that can recognize BECT patients. On the other hand, most of the existing approaches extract hand‐crafted features from MRI or fMRI, which cannot obtain the desired performance due to the limited representative capacity of the used features. To address these issues, we propose a novel classification method by fusing the predictions of three different views: hand‐crafted features view, MRI view, and fMRI view. The final result is obtained by passing through those predictions after a fusing neural network. The basic idea of our method is that multiple views could provide complementary information and thus can boost the classification performance. Extensive experiments show that the proposed multi‐view method is remarkably superior to single‐view methods
Intelligent gradient amplification for deep neural networks
Get PDFDeep learning models offer superior performance compared to other machine
learning techniques for a variety of tasks and domains, but pose their own
challenges. In particular, deep learning models require larger training times
as the depth of a model increases, and suffer from vanishing gradients. Several
solutions address these problems independently, but there have been minimal
efforts to identify an integrated solution that improves the performance of a
model by addressing vanishing gradients, as well as accelerates the training
process to achieve higher performance at larger learning rates. In this work,
we intelligently determine which layers of a deep learning model to apply
gradient amplification to, using a formulated approach that analyzes gradient
fluctuations of layers during training. Detailed experiments are performed for
simpler and deeper neural networks using two different intelligent measures and
two different thresholds that determine the amplification layers, and a
training strategy where gradients are amplified only during certain epochs.
Results show that our amplification offers better performance compared to the
original models, and achieves accuracy improvement of around 2.5% on CIFAR- 10
and around 4.5% on CIFAR-100 datasets, even when the models are trained with
higher learning rates
Using Transfer Learning, SVM, and Ensemble Classification to Classify Baby Cries Based on Their Spectrogram Images
No full text
