scholarly journals SVM-Based Deep Stacking Networks

2019 ◽  
Vol 33 ◽  
pp. 5273-5280 ◽  
Author(s):  
Jingyuan Wang ◽  
Kai Feng ◽  
Junjie Wu
Keyword(s):  
Neural Networks ◽  
High Performance ◽  
Optimal Solution ◽  
Data Sets ◽  
Complex Data ◽  
Layer By Layer ◽  
Deep Network ◽  
Local View ◽  
Linear Svm ◽  
Network Approaches

The deep network model, with the majority built on neural networks, has been proved to be a powerful framework to represent complex data for high performance machine learning. In recent years, more and more studies turn to nonneural network approaches to build diverse deep structures, and the Deep Stacking Network (DSN) model is one of such approaches that uses stacked easy-to-learn blocks to build a parameter-training-parallelizable deep network. In this paper, we propose a novel SVM-based Deep Stacking Network (SVM-DSN), which uses the DSN architecture to organize linear SVM classifiers for deep learning. A BP-like layer tuning scheme is also proposed to ensure holistic and local optimizations of stacked SVMs simultaneously. Some good math properties of SVM, such as the convex optimization, is introduced into the DSN framework by our model. From a global view, SVM-DSN can iteratively extract data representations layer by layer as a deep neural network but with parallelizability, and from a local view, each stacked SVM can converge to its optimal solution and obtain the support vectors, which compared with neural networks could lead to interesting improvements in anti-saturation and interpretability. Experimental results on both image and text data sets demonstrate the excellent performances of SVM-DSN compared with some competitive benchmark models.

Science Communication with Dinosaurs

2012 ◽  
pp. 587-611
Author(s):  
Phillip L. Manning ◽  
Peter L. Falkingham
Keyword(s):  
High Performance Computing ◽  
Public Engagement ◽  
Science Communication ◽  
High Performance ◽  
Computational Science ◽  
Data Sets ◽  
Complex Data ◽  
Multidisciplinary Research ◽  
Complex Data Sets ◽  
Performance Computing

Dinosaurs successfully conjure images of lost worlds and forgotten lives. Our understanding of these iconic, extinct animals now comes from many disciplines, not just the science of palaeontology. In recent years palaeontology has benefited from the application of new and existing techniques from physics, biology, chemistry, engineering, but especially computational science. The application of computers in palaeontology is highlighted in this chapter as a key area of development in studying fossils. The advances in high performance computing (HPC) have greatly aided and abetted multiple disciplines and technologies that are now feeding paleontological research, especially when dealing with large and complex data sets. We also give examples of how such multidisciplinary research can be used to communicate not only specific discoveries in palaeontology, but also the methods and ideas, from interrelated disciplines to wider audiences. Dinosaurs represent a useful vehicle that can help enable wider public engagement, communicating complex science in digestible chunks.

scholarly journals Optical Compute Engine Using Deep CNN

2018 ◽  
Vol 7 (2.24) ◽  
pp. 541
Author(s):  
Zainab Zaveri ◽  
Dhruv Gosain ◽  
Arul Prakash M
Keyword(s):  
Neural Networks ◽  
Error Correction ◽  
Image Classification ◽  
Layered Structure ◽  
Data Sets ◽  
Complex Data ◽  
Complex Data Sets ◽  
Deep Cnn ◽  
Classification Tasks ◽  
Hierarchical Manner

We present an optical compute engine with implementation of Deep CNNs. CNNs are designed in an organized and hierarchical manner and their convolutional layers, subsampling layers alternate with each other, thus   the intricacy of the data per layer escalates as we traverse in the layered structure, which gives us more efficient results when dealing with complex data sets and computations. CNNs are realised in a distinctive way and vary from other neural networks in how their convolutional and subsampling layers are organised. DCNNs bring us very proficient results when it comes to image classification tasks. Recently, we have understood that generalization is more important when compared to the neural network’s depth for more optimised image classification. Our feature extractors are learned in an unsupervised way, hence the results get more precise after every backpropagation and error correction.

scholarly journals Hyper Spectral Image Classification using Multi Labelled, Multi-Scale and Multi-Angle CNN with MS-MA BT Algorithm

2019 ◽  
Vol 8 (9S3) ◽  
pp. 1229-1234
Keyword(s):  
Neural Networks ◽  
High Performance ◽  
Hyperspectral Image ◽  
Support Vector ◽  
Data Sets ◽  
Feature Maps ◽  
Feature Map ◽  
Convolution Neural Networks ◽  
Multi Scale ◽  
Suggested Technique

For classifying the hyperspectral image (HSI), convolution neural networks are used widely as it gives high performance and better results. For stronger prediction this paper presents new structure that benefit from both MS - MA BT (multi-scale multi-angle breaking ties) and CNN algorithm. We build a new MS - MA BT and CNN architecture. It obtains multiple characteristics from the raw image as an input. This algorithm generates relevant feature maps which are fed into concatenating layer to form combined feature map. The obtained mixed feature map is then placed into the subsequent stages to estimate the final results for each hyperspectral pixel. Not only does the suggested technique benefit from improved extraction of characteristics from CNNs and MS-MA BT, but it also allows complete combined use of visual and temporal data. The performance of the suggested technique is evaluated using SAR data sets, and the results indicate that the MS-MA BT-based multi-functional training algorithm considerably increases identification precision. Recently, convolution neural networks have proved outstanding efficiency on multiple visual activities, including the ranking of common two-dimensional pictures. In this paper, the MS-MA BT multi-scale multi-angle CNN algorithm is used to identify hyperspectral images explicitly in the visual domain. Experimental outcomes based on several SAR image data sets show that the suggested technique can attain greater classification efficiency than some traditional techniques, such as support vector machines and conventional deep learning techniques.

scholarly journals BIG GEO DATA MANAGEMENT: AN EXPLORATION WITH SOCIAL MEDIA AND TELECOMMUNICATIONS OPEN DATA

2016 ◽  
Vol XLI-B2 ◽  
pp. 595-601
Author(s):  
C. Arias Munoz ◽  
M. A. Brovelli ◽  
S. Corti ◽  
G. Zamboni
Keyword(s):  
Data Integration ◽  
High Speed ◽  
High Performance ◽  
Open Data ◽  
Geospatial Data ◽  
Data Sets ◽  
Complex Data ◽  
Complex Data Sets ◽  
Computationally Intensive ◽  
The Government

The term Big Data has been recently used to define big, highly varied, complex data sets, which are created and updated at a high speed and require faster processing, namely, a reduced time to filter and analyse relevant data. These data is also increasingly becoming Open Data (data that can be freely distributed) made public by the government, agencies, private enterprises and among others. There are at least two issues that can obstruct the availability and use of Open Big Datasets: Firstly, the gathering and geoprocessing of these datasets are very computationally intensive; hence, it is necessary to integrate high-performance solutions, preferably internet based, to achieve the goals. Secondly, the problems of heterogeneity and inconsistency in geospatial data are well known and affect the data integration process, but is particularly problematic for Big Geo Data. Therefore, Big Geo Data integration will be one of the most challenging issues to solve. With these applications, we demonstrate that is possible to provide processed Big Geo Data to common users, using open geospatial standards and technologies. NoSQL databases like MongoDB and frameworks like RASDAMAN could offer different functionalities that facilitate working with larger volumes and more heterogeneous geospatial data sources.

scholarly journals Detection of Melanoma Skin Cancer with Deep Neural Networks

2019 ◽  
Vol 4 (4) ◽  
Keyword(s):  
Neural Networks ◽  
Skin Cancer ◽  
High Performance ◽  
Deep Neural Networks ◽  
Histopathological Examination ◽  
Skin Lesions ◽  
Image Features ◽  
Test Accuracy ◽  
Data Sets ◽  
Deep Convolutional Neural Networks

Detection of skin cancer involves several steps of examinations first being visual diagnosis that is followed by dermoscopic analysis, a biopsy, and histopathological examination. The classification of skin lesions in the first step is critical and challenging as classes vary by minute appearance in skin lesions. Deep convolutional neural networks (CNNs) have great potential in multicategory image-based classification by considering coarse-to-fine image features. This study aims to demonstrate how to classify skin lesions, in particular, melanoma, using CNN trained on data sets with disease labels. We developed and trained our own CNN model using a subset of the images from International Skin Imaging Collaboration (ISIC) Dermoscopic Archive. To test the performance of the proposed model, we used a different subset of images from the same archive as the test set. Our model is trained to classify images into two categories: malignant melanoma and nevus and is shown to achieve excellent classification results with high test accuracy (91.16%) and high performance as measured by various metrics. Our study demonstrated the potential of using deep neural networks to assist early detection of melanoma and thereby improve the patient survival rate from this aggressive skin cancer.

scholarly journals Learning Anytime Predictions in Neural Networks via Adaptive Loss Balancing

2019 ◽  
Vol 33 ◽  
pp. 3812-3821 ◽  
Author(s):  
Hanzhang Hu ◽  
Debadeepta Dey ◽  
Martial Hebert ◽  
J. Andrew Bagnell
Keyword(s):  
Neural Networks ◽  
High Performance ◽  
Geometric Mean ◽  
Computational Cost ◽  
Data Sets ◽  
Adaptive Weights ◽  
Multiple Viewpoints ◽  
The Cost ◽  
Theoretical Considerations ◽  
Adaptive Weighted Sum

This work considers the trade-off between accuracy and testtime computational cost of deep neural networks (DNNs) via anytime predictions from auxiliary predictions. Specifically, we optimize auxiliary losses jointly in an adaptive weighted sum, where the weights are inversely proportional to average of each loss. Intuitively, this balances the losses to have the same scale. We demonstrate theoretical considerations that motivate this approach from multiple viewpoints, including connecting it to optimizing the geometric mean of the expectation of each loss, an objective that ignores the scale of losses. Experimentally, the adaptive weights induce more competitive anytime predictions on multiple recognition data-sets and models than non-adaptive approaches including weighing all losses equally. In particular, anytime neural networks (ANNs) can achieve the same accuracy faster using adaptive weights on a small network than using static constant weights on a large one. For problems with high performance saturation, we also show a sequence of exponentially deepening ANNs can achieve near-optimal anytime results at any budget, at the cost of a const fraction of extra computation.

scholarly journals BIG GEO DATA MANAGEMENT: AN EXPLORATION WITH SOCIAL MEDIA AND TELECOMMUNICATIONS OPEN DATA

2016 ◽  
Vol XLI-B2 ◽  
pp. 595-601
Author(s):  
C. Arias Munoz ◽  
M. A. Brovelli ◽  
S. Corti ◽  
G. Zamboni
Keyword(s):  
Data Integration ◽  
High Speed ◽  
High Performance ◽  
Open Data ◽  
Geospatial Data ◽  
Data Sets ◽  
Complex Data ◽  
Complex Data Sets ◽  
Computationally Intensive ◽  
The Government

The term Big Data has been recently used to define big, highly varied, complex data sets, which are created and updated at a high speed and require faster processing, namely, a reduced time to filter and analyse relevant data. These data is also increasingly becoming Open Data (data that can be freely distributed) made public by the government, agencies, private enterprises and among others. There are at least two issues that can obstruct the availability and use of Open Big Datasets: Firstly, the gathering and geoprocessing of these datasets are very computationally intensive; hence, it is necessary to integrate high-performance solutions, preferably internet based, to achieve the goals. Secondly, the problems of heterogeneity and inconsistency in geospatial data are well known and affect the data integration process, but is particularly problematic for Big Geo Data. Therefore, Big Geo Data integration will be one of the most challenging issues to solve. With these applications, we demonstrate that is possible to provide processed Big Geo Data to common users, using open geospatial standards and technologies. NoSQL databases like MongoDB and frameworks like RASDAMAN could offer different functionalities that facilitate working with larger volumes and more heterogeneous geospatial data sources.

scholarly journals Examining the Causal Structures of Deep Neural Networks Using Information Theory

Entropy ◽  
2020 ◽  
Vol 22 (12) ◽  
pp. 1429
Author(s):  
Scythia Marrow ◽  
Eric J. Michaud ◽  
Erik Hoel
Keyword(s):  
Neural Networks ◽  
Information Theory ◽  
Mutual Information ◽  
Deep Neural Networks ◽  
Causal Structure ◽  
Data Sets ◽  
Layer By Layer ◽  
Input And Output ◽  
Causal Structures ◽  
Entropy Perturbation

Deep Neural Networks (DNNs) are often examined at the level of their response to input, such as analyzing the mutual information between nodes and data sets. Yet DNNs can also be examined at the level of causation, exploring “what does what” within the layers of the network itself. Historically, analyzing the causal structure of DNNs has received less attention than understanding their responses to input. Yet definitionally, generalizability must be a function of a DNN’s causal structure as it reflects how the DNN responds to unseen or even not-yet-defined future inputs. Here, we introduce a suite of metrics based on information theory to quantify and track changes in the causal structure of DNNs during training. Specifically, we introduce the effective information (EI) of a feedforward DNN, which is the mutual information between layer input and output following a maximum-entropy perturbation. The EI can be used to assess the degree of causal influence nodes and edges have over their downstream targets in each layer. We show that the EI can be further decomposed in order to examine the sensitivity of a layer (measured by how well edges transmit perturbations) and the degeneracy of a layer (measured by how edge overlap interferes with transmission), along with estimates of the amount of integrated information of a layer. Together, these properties define where each layer lies in the “causal plane”, which can be used to visualize how layer connectivity becomes more sensitive or degenerate over time, and how integration changes during training, revealing how the layer-by-layer causal structure differentiates. These results may help in understanding the generalization capabilities of DNNs and provide foundational tools for making DNNs both more generalizable and more explainable.

scholarly journals Neutron separation energies from nuclear mass systematics using neural networks

2019 ◽  
Vol 11 ◽  
Author(s):  
S. Athanassopoulos ◽  
E. Mavrommatis ◽  
K. A. Gernoth ◽  
J. W. Clark
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
High Performance ◽  
Ion Beam ◽  
Heavy Ion ◽  
Neural Network Modeling ◽  
Data Sets ◽  
Global Models ◽  
Coding Schemes ◽  
The Masses

Statistical modeling of data sets by neural-network techniques is offered as an alternative to traditional semi-empirical approaches to global modeling of nuclear properties. There is need for such systematics driven by fundamental investigations of nuclear structure far from stability, conducted at heavy-ion and radioactive-ion beam facilities. There is also great current interest from the perspective of astrophysics and of nuclear technology. In this work we evaluate the one and two neutron separation energies based on global models for the masses of nuclides developed with the use of neural networks[l] and compare them with the experimental ones as given by Audi at Atomic Mass Data Center web site [2], Our work on masses is a continuation of the work reported in ref. [3]. We have used enriched data sets together with a novel training algorithm and various coding schemes to achieve high performance both in learning and prediction. Our performance is comparable to the best of other evaluations of separation energies based on global models for the masses of nuclides (like those of Möller et al. [4] and Pearson et al. [5] that are rooted in conventional Hamiltonian theory), whereas the number of parameters is larger. Neural network modeling, as well as other statistical strategies based on new algorithms for artificial intelligence, may prove to be a useful asset in the further exploration of nuclear phenomena far from stability.

scholarly journals Crystal Engineering Approach for Fabrication of Inverted Perovskite Solar Cell in Ambient Conditions

Energies ◽  
2021 ◽  
Vol 14 (6) ◽  
pp. 1751
Author(s):  
Inga Ermanova ◽  
Narges Yaghoobi Nia ◽  
Enrico Lamanna ◽  
Elisabetta Di Bartolomeo ◽  
Evgeny Kolesnikov ◽  
...  
Keyword(s):  
Solar Cell ◽  
Crystal Engineering ◽  
High Performance ◽  
Perovskite Solar Cells ◽  
Perovskite Solar Cell ◽  
Layer By Layer ◽  
Ambient Conditions ◽  
Sequential Method ◽  
Engineering Approach ◽  
Hole Transporting

In this paper, we demonstrate the high potentialities of pristine single-cation and mixed cation/anion perovskite solar cells (PSC) fabricated by sequential method deposition in p-i-n planar architecture (ITO/NiOX/Perovskite/PCBM/BCP/Ag) in ambient conditions. We applied the crystal engineering approach for perovskite deposition to control the quality and crystallinity of the light-harvesting film. The formation of a full converted and uniform perovskite absorber layer from poriferous pre-film on a planar hole transporting layer (HTL) is one of the crucial factors for the fabrication of high-performance PSCs. We show that the in-air sequential deposited MAPbI3-based PSCs on planar nickel oxide (NiOX) permitted to obtain a Power Conversion Efficiency (PCE) exceeding 14% while the (FA,MA,Cs)Pb(I,Br)3-based PSC achieved 15.6%. In this paper we also compared the influence of transporting layers on the cell performance by testing material depositions quantity and thickness (for hole transporting layer), and conditions of deposition processes (for electron transporting layer). Moreover, we optimized second step of perovskite deposition by varying the dipping time of substrates into the MA(I,Br) solution. We have shown that the layer by layer deposition of the NiOx is the key point to improve the efficiency for inverted perovskite solar cell out of glove-box using sequential deposition method, increasing the relative efficiency of +26% with respect to reference cells.

Sign in / Sign up

Export Citation Format

Share Document