Conventional Two-stage Method Research Articles

NLGC: Network localized Granger causality with application to MEG directional functional connectivity analysis

Identifying the directed connectivity that underlie networked activity between different cortical areas is critical for understanding the neural mechanisms behind sensory processing. Granger causality (GC) is widely used for this purpose in functional magnetic resonance imaging analysis, but there the temporal resolution is low, making it difficult to capture the millisecond-scale interactions underlying sensory processing. Magnetoencephalography (MEG) has millisecond resolution, but only provides low-dimensional sensor-level linear mixtures of neural sources, which makes GC inference challenging. Conventional methods proceed in two stages: First, cortical sources are estimated from MEG using a source localization technique, followed by GC inference among the estimated sources. However, the spatiotemporal biases in estimating sources propagate into the subsequent GC analysis stage, may result in both false alarms and missing true GC links. Here, we introduce the Network Localized Granger Causality (NLGC) inference paradigm, which models the source dynamics as latent sparse multivariate autoregressive processes and estimates their parameters directly from the MEG measurements, integrated with source localization, and employs the resulting parameter estimates to produce a precise statistical characterization of the detected GC links. We offer several theoretical and algorithmic innovations within NLGC and further examine its utility via comprehensive simulations and application to MEG data from an auditory task involving tone processing from both younger and older participants. Our simulation studies reveal that NLGC is markedly robust with respect to model mismatch, network size, and low signal-to-noise ratio, whereas the conventional two-stage methods result in high false alarms and mis-detections. We also demonstrate the advantages of NLGC in revealing the cortical network-level characterization of neural activity during tone processing and resting state by delineating task- and age-related connectivity changes.

DIRV: Dense Interaction Region Voting for End-to-End Human-Object Interaction Detection

Recent years, human-object interaction (HOI) detection has achieved impressive advances. However, conventional two-stage methods are usually slow in inference. On the other hand, existing one-stage methods mainly focus on the union regions of interactions, which introduce unnecessary visual information as disturbances to HOI detection. To tackle the problems above, we propose a novel one-stage HOI detection approach DIRV in this paper, based on a new concept called interaction region for the HOI problem. Unlike previous methods, our approach concentrates on the densely sampled interaction regions across different scales for each human-object pair, so as to capture the subtle visual features that is most essential to the interaction. Moreover, in order to compensate for the detection flaws of a single interaction region, we introduce a novel voting strategy that makes full use of those overlapped interaction regions in place of conventional Non-Maximal Suppression (NMS). Extensive experiments on two popular benchmarks: V-COCO and HICO-DET show that our approach outperforms existing state-of-the-arts by a large margin with the highest inference speed and lightest network architecture. Our code is publicly available at www.github.com/MVIG-SJTU/DIRV.

A Question Answering-Based Framework for One-Step Event Argument Extraction

Event argument extraction, which aims to identify arguments of specific events and label their roles, is a challenging subtask of event extraction. Previous approaches solve this problem in a two-stage manner that first extracts named entities as argument candidates and then determines their roles. However, many nested entities may be missed or wrongly predicted during the argument candidate extraction procedure, which substantially affects the performance of the downstream classifier. In this paper, we propose a novel one-step question answering based framework, which performs argument candidate extraction and argument role classification simultaneously to mitigate the error propagation problem in conventional two-stage methods. Since the conventional question answering based framework cannot be applied directly to this task, we design a Q uestion A nswering based S equence L abeling (QA-SL) model to tackle inexistent argument roles and multiple argument token spans. Moreover, considering the overwhelming number of parameters in question answering based neural network models and the relatively small size of event extraction corpus, we fine-tune the pre-trained model from BERT to mitigate the data scarcity problem. Extensive experiments demonstrate the benefits of the proposed method, leading to a competitive performance compared with state-of-the-art methods. To the best of our knowledge, this is the first work to cast event argument extraction as a question answering task.

Optimal reconfiguration of active distribution network considering flexible load scheduling

At present, most of the existing optimal reconfiguration method of active distribution network (ADN) aims at the change of network topology, without considering flexible load. This situation makes a low efficiency of the existing method when facing the situation with large amount of flexible load in ADN. A bi-layer optimal reconfiguration model including an outer layer and inner layer is presented considering flexible load optimal dispatching to improve the efficiency of ADN reconfiguration. Unlike the conventional method, the two layers of the proposed bi-layer method are combined in each iteration. At each iteration the reconfiguration scheme is obtained by solving a binary optimal model in the outer layer, and the flexible loads are optimized in the inner layer. So, the optimal network topology and the optimal flexible load scheme are obtained at the same iteration. This would increase the efficiency of the reconfiguration scheme than that get by the conventional two-stage method. Finally, the importance of flexible load during ADN reconfiguration is proved, and the validity of the proposed method is verified by the comparison simulations.

Deep Binary Reconstruction for Cross-Modal Hashing

To satisfy the huge storage space and organization capacity requirements in addressing big multimodal data, hashing techniques have been widely employed to learn binary representations in cross-modal retrieval tasks. However, optimizing the hashing objective under the necessary binary constraint is truly a difficult problem. A common strategy is to relax the constraint and perform individual binarizations over the learned real-valued representations. In this paper, in contrast to conventional two-stage methods, we propose to directly learn the binary codes, where the model can be easily optimized by a standard gradient descent optimizer. However, before that, we present a theoretical guarantee of the effectiveness of the multimodal network in preserving the inter- and intra-modal consistencies. Based on this guarantee, a novel multimodal deep binary reconstruction model is proposed, which can be trained to simultaneously model the correlation across modalities and learn the binary hashing codes. To generate binary codes and to avoid the tiny gradient problem, a novel activation function first scales the input activations to suitable scopes and, then, feeds them to the tanh function to build the hashing layer. Such a composite function is named adaptive tanh . Both linear and nonlinear scaling methods are proposed and shown to generate efficient codes after training the network. Extensive ablation studies and comparison experiments are conducted for the image2text and text2image retrieval tasks; the method is found to outperform several state-of-the-art deep-learning methods with respect to different evaluation metrics.

Blocking of matrix metalloproteinases-13 responsive peptide in poly(urethane urea) for potential cartilage tissue engineering applications.

The matching of scaffold degradation rate with neotissue growth is required for tissue engineering applications. Timely provision of proper spaces especially for cartilage tissue engineering plays a pivotal role in chondrocyte cluster formation. In this study, poly(urethane urea) was synthesized using conventional two-stage method by extending the isocyanate group terminated prepolymers with different amounts of GPLGLWARK peptide, which responses the degrading induced by matrix metalloproteinase 13, the main proteinase for cartilage matrix degradation. The Fourier transform infrared spectrometer with the attenuated total reflection and 1H nuclear magnetic resonance spectra revealed that the peptides were introduced to poly(urethane urea) according to the characteristic absorption bands of the peptide and the newly formed urea bonds. The ultraviolet-visible spectroscopy spectra showed that the weight percentages of the peptide in the three poly(urethane urea) were 25%, 32%, and 35%. Atomic force microscopy images revealed that phase separation occurred in all poly(urethane urea) samples and became increasingly apparent with increasing amount of peptides introduced. Mechanical tests showed that the poly(urethane urea) strength increased with increasing amount of peptides in poly(urethane urea). Poly(urethane urea) proteolysis in matrix metalloproteinase 13 solution was more rapid than hydrolysis in aqueous buffer, and proteolysis rate was dependent on the amount of peptides in poly(urethane urea). Cell proliferation on the material surface in vitro displayed nontoxicity for all synthesized poly(urethane urea). In vivo subcutaneous implantation evaluation revealed the presence of local foreign body reactions triggered by poly(urethane urea) but was not due to peptide in poly(urethane urea). Moreover, the synthesized poly(urethane urea) with significant phase separation did not degrade under the matrix metalloproteinase 13 free subcutaneous environment, but poly(urethane urea) with minimal phase separation was degraded by attacking of the enzymes adsorbed on the hydrophobic surface through non-specific adsorption.

Unifying Blind Separation and Clustering for Resting-State EEG/MEG Functional Connectivity Analysis

Unsupervised analysis of the dynamics (nonstationarity) of functional brain connectivity during rest has recently received a lot of attention in the neuroimaging and neuroengineering communities. Most studies have used functional magnetic resonance imaging, but electroencephalography (EEG) and magnetoencephalography (MEG) also hold great promise for analyzing nonstationary functional connectivity with high temporal resolution. Previous EEG/MEG analyses divided the problem into two consecutive stages: the separation of neural sources and then the connectivity analysis of the separated sources. Such nonoptimal division into two stages may bias the result because of the different prior assumptions made about the data in the two stages. We propose a unified method for separating EEG/MEG sources and learning their functional connectivity (coactivation) patterns. We combine blind source separation (BSS) with unsupervised clustering of the activity levels of the sources in a single probabilistic model. A BSS is performed on the Hilbert transforms of band-limited EEG/MEG signals, and coactivation patterns are learned by a mixture model of source envelopes. Simulation studies show that the unified approach often outperforms conventional two-stage methods, indicating further the benefit of using Hilbert transforms to deal with oscillatory sources. Experiments on resting-state EEG data, acquired in conjunction with a cued motor imagery or nonimagery task, also show that the states (clusters) obtained by the proposed method often correlate better with physiologically meaningful quantities than those obtained by a two-stage method.

Unifying blind separation and clustering for resting-state EEG/MEG functional connectivity analysis

Event Abstract Back to Event Unifying blind separation and clustering for resting-state EEG/MEG functional connectivity analysis Jun-ichiro Hirayama1*, Takeshi Ogawa2 and Aapo Hyvärinen3 1 ATR, Neural Information Analysis Laboratories, Japan 2 ATR, Cognitive Mechanisms Laboratories, Japan 3 University of Helsinki, Department of Computer Science and HIIT, Finland Unsupervised analysis of non-stationary functional brain connectivity during rest has recently received a lot of attention in the field of functional brain imaging. Most studies have used functional magnetic resonance imaging, but electroencephalography (EEG) and magnetoencephalography (MEG) also hold great promise for analyzing nonstationary functional connectivity with high temporal resolution. However, conventional two-stage analyses of resting-state EEG/MEG may be nonoptimal as they divide the problem into two stages making different prior assumptions about the data, i.e., separation of neural sources and subsequent connectivity analysis of the separated sources. Here, we propose a first unifying approach to separating EEG/MEG sources and learning their functional connectivity (coactivation) patterns simultaneously. Specifically, we combine blind source separation (BSS) with unsupervised clustering of the activity levels of the sources in a single probabilistic model. The BSS is performed on the Hilbert transforms of band-limited EEG/MEG signals, and coactivation patterns are learned by a mixture model of source envelopes. Simulation studies show that the unified approach often outperforms conventional two-stage methods, indicating further the benefit of using Hilbert transform to deal with oscillatory sources. Experiments on resting-state EEG data, acquired in conjunction with a cued motor imagery or nonimagery task, also show that the states (clusters) obtained by the proposed method are often more interpretable than those obtained by a two-stage method. Acknowledgements This work was supported by a contract with the Ministry of Internal Affairs and Communications “Novel and Innovative R&D Making Use of Brain Structures” and by JSPS KAKENHI Grant Number 25730155. J.H. was partially supported by Strategic International Research Cooperative Program, Japan Science and Technology Agency (JST). A.H. was supported by the Academy of Finland, Centre-of-Excellence in Inverse Problems Research and Centre-of-Excellence Algorithmic Data Analysis. References J. Hirayama, T. Ogawa and A. Hyvarinen. Unifying blind separation and clustering for resting-state EEG/MEG functional connectivity analysis. Neural Computation, 27(7), 1373-1404, 2015. Keywords: unsupervised learning, functional connectivity, resting state, Electroencephalography (EEG), Magnetoencephalography (MEG) Conference: German-Japanese Adaptive BCI Workshop, Kyoto, Japan, 28 Oct - 29 Oct, 2015. Presentation Type: Oral presentation (Invited speakers) Topic: Adaptive BCI Citation: Hirayama J, Ogawa T and Hyvärinen A (2015). Unifying blind separation and clustering for resting-state EEG/MEG functional connectivity analysis. Front. Comput. Neurosci. Conference Abstract: German-Japanese Adaptive BCI Workshop. doi: 10.3389/conf.fncom.2015.56.00010 Copyright: The abstracts in this collection have not been subject to any Frontiers peer review or checks, and are not endorsed by Frontiers. They are made available through the Frontiers publishing platform as a service to conference organizers and presenters. The copyright in the individual abstracts is owned by the author of each abstract or his/her employer unless otherwise stated. Each abstract, as well as the collection of abstracts, are published under a Creative Commons CC-BY 4.0 (attribution) licence (https://creativecommons.org/licenses/by/4.0/) and may thus be reproduced, translated, adapted and be the subject of derivative works provided the authors and Frontiers are attributed. For Frontiers’ terms and conditions please see https://www.frontiersin.org/legal/terms-and-conditions. Received: 15 Oct 2015; Published Online: 04 Nov 2015. * Correspondence: Dr. Jun-ichiro Hirayama, ATR, Neural Information Analysis Laboratories, Seika-cho, Soraku-gun, Kyoto, 6190288, Japan, hirayama@atr.jp Login Required This action requires you to be registered with Frontiers and logged in. To register or login click here. Abstract Info Abstract The Authors in Frontiers Jun-ichiro Hirayama Takeshi Ogawa Aapo Hyvärinen Google Jun-ichiro Hirayama Takeshi Ogawa Aapo Hyvärinen Google Scholar Jun-ichiro Hirayama Takeshi Ogawa Aapo Hyvärinen PubMed Jun-ichiro Hirayama Takeshi Ogawa Aapo Hyvärinen Related Article in Frontiers Google Scholar PubMed Abstract Close Back to top Javascript is disabled. Please enable Javascript in your browser settings in order to see all the content on this page.

Synthesis and characterization of degradable polyimides derived from 4,6-bis(4-aminophenoxy)pyrimidine

Polyimides (PIs) are widely applied in the fields of aerospace, membrane, and electronic industries. However, with the fast-growing application, the degradation of polymer has become a serious problem people are confronted with. In this work, a series of novel degradable PIs were prepared from diamine 4,6-bis(4-aminophenoxy)pyrimidine with various commercially available aromatic dianhydrides by a conventional two-stage method. The resulting PIs could degrade fast under methylamine medium in N,N-dimethylacetamide, and the degradation mechanism was speculated by analyzing the degradation products. Compared with the traditional method, a facile and effective method for PI degradation was provided. The results suggested that incorporating 4,6-diphenoxypyrimidine structure into polymer chains may grow up to be a new method for polymer degradation.

Characterization and Thermal Degradation of Polyimide Derived from 5-amino-2(p-aminophenyl) Benzoxazole Monomer with Pyromellitic Dianhydride

- 5-amino-2(p-aminophenyl) benzoxazole (AAPB) and its derived polyimide prepared with pyromellitic dianhydride (PMDA) were synthesized successfully. Their structures were characterized by FT-IR, 1H NMR, Element analysis, 13C NMR, Solid-state 13CNMR, DSC, TG and XRD. The polyimide was synthesized via a conventional two-stage method. The intermediate poly (amic acid) had inherent viscosity of 1.06dL/g and could be thermally converted into light yellow polyimide film. The polyimide showed excellent solvent resistance and good thermal stability. The glass transition temperature (Tg) was found to be 314°C, the decomposition started at a temperature above 500°C in air and above 550°C in argon atmosphere. The thermal degradation of the polyimide was studied by thermogravimetric analysis (TGA) in order to determine the actual reaction mechanisms of the decomposition process. The apparent activation energy (Ea) was obtained following Flynn-Wall-Ozawa method. The activation energies of different mechanism models and pre-exponential factor (A) were determined by Coats-Redfern method. Compared with the value obtained from the Ozawa method, the actual reaction mechanism obeyed three-dimensional diffusion model, Jander equation (D3) with integral form g(α)=[1-(1-α)1/3]2.

Characterization and Thermal Degradation of Polyimide Derived from 5-amino-2(p-aminophenyl) Benzoxazole Monomer with Pyromellitic Dianhydride

- 5-amino-2(p-aminophenyl) benzoxazole (AAPB) and its derived polyimide prepared with pyromellitic dianhydride (PMDA) were synthesized successfully. Their structures were characterized by FT-IR, 1H NMR, Element analysis, 13C NMR, Solid-state 13CNMR, DSC, TG and XRD. The polyimide was synthesized via a conventional two-stage method. The intermediate poly (amic acid) had inherent viscosity of 1.06dL/g and could be thermally converted into light yellow polyimide film. The polyimide showed excellent solvent resistance and good thermal stability. The glass transition temperature (Tg) was found to be 314°C, the decomposition started at a temperature above 500°C in air and above 550°C in argon atmosphere. The thermal degradation of the polyimide was studied by thermogravimetric analysis (TGA) in order to determine the actual reaction mechanisms of the decomposition process. The apparent activation energy (Ea) was obtained following Flynn-Wall-Ozawa method. The activation energies of different mechanism models and pre-exponential factor (A) were determined by Coats-Redfern method. Compared with the value obtained from the Ozawa method, the actual reaction mechanism obeyed three-dimensional diffusion model, Jander equation (D3) with integral form g(α)=[1-(1-α)1/3]2.

Synthesis and Characterization of Polyimides Derived from 5-Amino-2(p-aminophenyl) Benzoxazole Monomer

Polyimides prepared from 5-amino-2(p-aminophenyl) benzoxazole (AAPB) with pyromellitic dianhydride(PMDA), 2,2’3,3’-biphenyl tetracarboxylic dianhydride(i-BPDA), 4,4’-oxdiphthalic anhydride(ODPA), 1,4-bis(3,4-dicarbosyphenoxy) benzene dianhydride(HQDPA) and 2,2-bis(3,4-dicarboxyphenyl)hexafluoropropane dianhydride (6FDA) were synthesized successfully via a conventional two-stage method. Their structures and properties were characterized by FT-IR, DSC, TG and XRD. The intermediate poly(amic acid)s had inherent viscosity range of 0.86~2.18 dL/g and could be thermally converted into light yellow polyimide films. The polyimides showed excellent solvent resistance and good thermal stability. The glass transition temperatures (Tg) were found to be 336oC~369 oC .The decomposition started at a temperature above 530oC in N2 atmosphere.

Bias and precision in the DEA two-stage method

In Data Envelopment Analysis (DEA), the two-stage method is a popular procedure for accounting for exogenous influences on efficiency. With the conventional two-stage method, a DEA is first conducted using only traditional (endogenous) inputs and outputs. Then, the first-stage DEA scores are regressed on the environmental/contextual (exogenous) inputs of interest. The regression outcomes are used to identify exogenous inputs that influence the first-stage DEA scores to a statistically significant degree, and to adjust DEA scores to account for these influences. Herein, it is demonstrated empirically that the conventional method exhibits substantial bias and low precision, with the degree of bias and precision affected by input variance and correlation. A reverse two-stage procedure that yields estimates without the bias and precision problems that compromise the validity of the conventional method's estimates is suggested.

Synthesis and Properties of New Aromatic Polyamides and Polyimides based on 1,4-Bis[N-(4-aminobenzoyl)-N-phenyl]phenylenediamine

A new N-phenylated amide (N-phenylamide) unit-containing aromatic diamine, 1,4-bis[N-(4-aminobenzoyl)-N-phenyl]phenylenediamine, was prepared by the condensation of NŃ-diphenyl-1,4-phenylenediamine with 4-nitrobenzoyl chloride, followed by catalytic reduction. Two series of aromatic poly(N-phenylamide-imide)s and poly(N-phenylamide-amide)s with inherent viscosities of 0.25-0.42 and 0.25-0.54 dL g-1 were prepared by a conventional two-stage method and the direct phosphorylation polycondensation, respectively, from the diamine with various aromatic dianhydrides and aromatic dicarboxylic acids. Some polymers based on rigid dianhydrides and diacids were semi-crystalline and insoluble in organic solvents. The other polymers were amorphous and organosoluble and could afford freestanding or tough films via solution casting. These polyimides and polyamides had glass transition temperatures in the range of 248-267 and 177-198°C, respectively. Decomposition temperatures of the polyimides for 10% weight loss all occurred above 470°C in both nitrogen and air atmospheres.

Integration of self-organizing feature map and K-means algorithm for market segmentation

Cluster analysis is a common tool for market segmentation. Conventional research usually employs the multivariate analysis procedures. In recent years, due to their high performance in engineering, artificial neural networks have also been applied in the area of management. Thus, this study aims to compare three clustering methods: (1) the conventional two-stage method, (2) the self-organizing feature maps and (3) our proposed two-stage method, via both simulated and real-world data. The proposed two-stage method is a combination of the self-organizing feature maps and the K-means method. The simulation results indicate that the proposed scheme is slightly better than the conventional two-stage method with respect to the rate of misclassification, and the real-world data on the basis of Wilk's Lambda and discriminant analysis. Scope and purpose The general idea of segmentation, or clustering, is to group items that are similar. A commonly used method is the multivariate analysis [4]. These methods consist of hierarchical methods, like Ward's minimum variance method, and the non-hierarchical methods, such as the K-means method. Owing to increase in computer power and decrease in computer costs, artificial neural networks (ANNs), which are distributed and parallel information processing systems successfully applied in the area of engineering, have recently been employed to solve the marketing problems. This study aims to discuss the possibility of integrating ANN and multivariate analysis. A two-stage method, which first uses the self-organizing feature maps to determine the number of clusters and the starting point and then employs the K-means method to find the final solution, is proposed. This method provides the marketing analysts a more sophisticated way to analyze the consumer behavior and determine the marking strategy. A case study is also employed to demonstrate the validity of the proposed method.

Cluster analysis in industrial market segmentation through artificial neural network

Market segmentation has commonly applied cluster analysis. This study intends to make the comparison of conventional two-stage method with proposed two-stage method through the simulated data. The proposed two-stage method is the combination of self-organizing feature maps and K-means method. The simulation results show that the proposed scheme is better than the conventional two-stage method based on the rate of misclassification.

Effects of Gasifying Agents on the Characterization of Nut Shell-derived Activated Carbon

Activated carbon was prepared from nut shells using a conventional two-stage method: carbonization followed by activation. Activation with steam or carbon dioxide as activating agent produced a range of chars of different burn-off. These were characterized for their total and micropore surface areas, and benzene adsorption capacity. Benzene adsorption measurement provided an insight into the effect of porosity development on the adsorptive properties of the adsorbent. It was found that activated carbon products from nut shells were comparable, in terms of adsorption characteristics, with activated carbons from other lignocellulosic precursors. The evolution of porosity of the resulting carbons shows that carbon dioxide is the preferable agent for the production of activated carbon with a narrow micropore size distribution.