Influence of Sequence Length and Geographic Representation on Optimal Prediction Architectures for Stolen Vehicle Geolocation

Optimal architecture of a convolutional neural network to estimate structural responses for safety evaluation of the structures

1. An experiment is described in which 96 individually-caged SCWL hens in two rooms were used to investigate the response to changing to a 28-h cycle using a reverse treatment design. 2. The application of the 28-h cycle did not affect mean rate of lay but increased mean egg weight and egg output. 3. Grouping the birds according to their preliminary sequence length yielded an interesting outcome. In both rooms, birds with short sequences (< or = 6 eggs) produced significantly more eggs under the 28-h cycle, while those with long sequences (> 6 eggs) produced marginally fewer eggs. The same trend was also evident with egg output. 4. Changing from 24-h to 28-h increased yolk and albumen weights as well as shell quality. However, relative to egg weight, no measurable effect was detected due to light cycle, age or sequence length on yolk and albumen weights. 5. The paper provides new evidence suggesting that long ahemeral cycles could be used to improve egg production as well as shell thickness in flocks with modest rates of lay.

The advent of rapid evolution on sequencing capacity of new genomes has evidenced the need for data analysis automation aiming at speeding up the genomic annotation process and reducing its cost. Given that one important step for functional genomic annotation is the promoter identification, several studies have been taken in order to propose computational approaches to predict promoters. Different classifiers and characteristics of the promoter sequences have been used to deal with this prediction problem. However, several works in literature have addressed the promoter prediction problem using datasets containing sequences of 250 nucleotides or more. As the sequence length defines the amount of dataset attributes, even considering a limited number of properties to characterize the sequences, datasets with a high number of attributes are generated for training classifiers. Once high-dimensional datasets can degrade the classifiers predictive performance or even require an infesible processing time, predicting promoters by training classifiers from datasets with a reduced number of attributes, it is essential to obtain good predictive performance with low computational cost. To the best of our knowledge, there is no work in literature that verified in a sistematic way the relation between the sequences length and the predictive performance of classifiers. Thus, in this work, sixteen datasets composed of different sized sequences are built and evaluated using the SVM and k-NN classifiers. The experimental results show that several datasets composed of shorter sequences acheived better predictive performance when compared with datasets composed of longer sequences and consumed a significantly shorter processing time.KeywordsPromoter SequenceSequence LengthPredictive PerformanceNucleosome PositionShort Processing TimeThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

The relation between gap discrimination and stream segregation [A. S. Bregman and I. Campbell, J. Exp. Psychol. 89, 244–249 (1971)] was examined using sequences of four 200‐ms, eight 100‐ms, or sixteen 50‐ms tones with a total frequency range range from 18 to 212 octaves. A gap of variable duration was inserted between tones of similar or disparate frequency. Gap discrimination, in a 2IFC procedure, became poorer as the difference between the frequencies of the tones bounding the gap increased, but was not influenced by changes in sequence length and rate. Judgments of stream segregation for the same stimuli by the same listeners showed strong individual differences, an influence of sequence length and rate as well as marker‐frequency difference, but little relation to gap discrimination. [Work supported by Grant NS‐14709 from NINCDS.]

In this article, we characterise the impact of product global performance on the choice of product architecture during the new product development (NPD) process. We classify product architectures into three categories: modular (neither interface is geometrically nested) integral (all interfaces are geometrically nested) and hybrid (some, but not all interfaces are geometrically nested). Existing studies show that the choice of a product architecture during the NPD process is a crucially strategic decision for a manufacturing firm. However, no single architecture is optimal in all cases, thus analytical models are required to identify and discuss specific trade-offs associated with the choice of the optimal architecture under different circumstances. This article develops analytical models that obtain a product's global performance through a modular/hybrid/integral architecture. Trade-offs between costs and expected benefits from different product architectures are analysed and compared. Multifunction products and small size are used as examples to formalise the models and show the impact of the global performance characteristics. We also investigate how optimal architecture changes in response to the exogenous costs of system integrators. Some empirical implications obtained from this study show that if the global performance benefit is sufficiently large, then modular architecture is an absolutely sub-optimal decision and integral architecture is an all-the-time candidate for optimal architecture.

Oracles programs accept information from various sources, transform it, and transmit it to smart contracts. They can also accept data from a smart contract and transmit it to an external data source. Ensuring the security, validity and integrity of the supplied data determines the success of the blockchain system, therefore, the research topic is relevant. The purpose of this article is to identify practically important features of Oracle programs and develop a version of the information system architecture for Oracles programs that meets the necessary requirements. The authors were faced with the task of investigating all the vulnerabilities associated with the use of Oracle programs and developing an optimal architectural solution. In the course of research, methods of reviewing scientific literature on the subject of research, collecting, structuring and analyzing the information received, and methods of choosing solutions were used. As a result of the research, the concept of an intelligent system for transferring external data to a blockchain management system is proposed and the optimal architecture of this intelligent system is developed. This solution is aimed at improving the security of using Oracle programs for blockchain management systems, especially blockchain management systems for industrial Internet of things applications. The solution can be used by developers of distributed registry systems to effectively launch and implement projects.

Because of the complexity and randomness of the gas-solid two-phase flow, it is still hard to fully understand the dynamic behavior of the flow. The electrostatic signal contains rich information of the particulate flow. Multiscale entropy (MSE) analysis of electrostatic fluctuation signal was conducted in this paper. The electrostatic signals were obtained from the horizontal pipeline in dilute pneumatic conveying system. The influence of sequence length on MSE of electrostatic signal was analyzed at first. Then the influence of solids concentration and superficial air velocity on MSE was shown, which reflects that it is helpful for understanding of the dynamics of two phase flow.

SWR is a member of chromatin remodeler family and participates the replacement of histone H2A with H2A.Z. One of the SWR subunits, Swc5, has an intrinsically disordered region and binds to H2A-H2B dimer. Though the binding structure of Swc5 and H2A-H2B has been resolved recently, it is still challenging to investigate the binding mechanism as well as the role of the charge interactions between Swc5 and H2A-H2B. Here we developed a coarse-grained structure-based model and performed molecular dynamics simulations to investigate the binding processes of two Swc5 regions with different lengths (swc5-a and swc5-b) to H2A-H2B. The simulation results suggest a different role of electrostatic interactions between swc5-a/swc5-b and H2A-H2B on binding. The electrostatic interactions between swc5-a/swc5-b and H2A-H2B can not only accelerate the initial capture step of binding, but can also trap the swc5-a/swc5-b at the wrong binding site on H2A. Besides, the conserved DEF/Y-2 motif of Swc5 is important for the binding affinity and the recognition with H2A-H2B at the initial step. Both swc5-a and swc5-b undergo a structural shift before reaching the final bound state. This theoretical study provides important details and the underlying physical mechanisms of the binding processes of swc5-a/swc5-b and H2A-H2B.

Study regionThe Ca River basin is located in the North Central Coast area of Vietnam Study focusThis study aims to develop a deep learning framework that is both effective and straightforward in order to forecast water levels in the Ca River basin in advance of multiple time steps for the event scales. We have thoroughly studied and assessed two deep learning models (DLMs), long-short term memory (LSTM) and gated recurrent unit (GRU), for their capacity to forecast water levels, focusing on various aspects such as the influence of sequence length or the impact of hyperparameter selection. Besides, two data scenarios were established using hydrological data from eight severe floods between 2007 and 2019 to examine the effect of input variables on model performance. Water level data was employed for both the scenarios (S1 and S2), whereas precipitation data was used only in S2. The cross-validation technique was used dynamically to address the issue of limited data. The inputs were reformatted as tensors and were then randomly divided into subsets. This flexible tuning preserved the sequential nature of the hydrological data while enabling the DLMs to be trained efficiently. New hydrological insights for the regionThe findings revealed that both the models exhibited equally excellent performances. The NSE of the LSTM model varies from 0.999–0.971 compared to 0.998–0.974 of the GRU model, corresponding to forecast cases from one to four-time steps ahead. This indicated that the use of multiple-input data types (S2) contrary to only one date type (S1) does not necessarily improve the forecasting performance. LSTM/GRU models with one hidden layer are adequate for delivering high performance while minimizing the data processing time.

Mobile devices have been playing significant roles in our daily lives, which has made device security and privacy protection extremely important. These mobile devices storing user sensitive and private information, therefore, need rigorous user authentication mechanisms. In this article, we present SearchAuth, a novel continuous authentication system on smartphones exploiting a neural architecture search (NAS) to find an optimal network architecture and an auto augmentation search (AAS) to more effectively train the optimal network along with the best data augmentation policies, by leveraging the accelerometer, gyroscope, and magnetometer on smartphones to capture users’ behavioral patterns. Specifically, SearchAuth consists of three stages, i.e., the offline stage, registration stage, and authentication stage. In the offline stage, we utilize the NAS on sensor data of the accelerometer, gyroscope, and magnetometer to find an optimal network architecture based on the designed search space. With the optimal network architecture, namely, NAS-based model, the AAS automatically optimizes the augmentation of the input data for more effectively training the model that is for feature extraction. In the registration stage, we use the trained NAS-based model to learn and extract deep features from the legitimate user’s data, and train the LOF classifier with 55 features selected by the PCA. In the authentication stage, with the well-trained NAS-based model and LOF classifier, SearchAuth identifies the current user as a legitimate user or an impostor when the user starts operating a smartphone. Based on our dataset, we evaluate the performance of the proposed SearchAuth, and the experimental results demonstrate that SearchAuth surpasses the representative authentication schemes by achieving the best accuracy of 93.95%, F1-score of 94.30%, and EER of 5.30% on the LOF classifier with dataset size of 100.

Deep neural networks enjoy high interest and have become the state-of-art methods in many fields of machine learning recently. Still, there is no easy way for a choice of network architecture. However, the choice of architecture can significantly influence the network performance. This work is the first step towards an automatic architecture design. We propose a genetic algorithm for an optimization of a network architecture. The algorithm is inspired by and designed directly for the Keras library [1] that is one of the most common implementations of deep neural networks. The target application is the prediction of air pollution based on sensor measurements. The proposed algorithm is evaluated on experiments on sensor data and compared to several fixed architectures and support vector regression.

In this chapter, we characterize the impact of product global performance on the choice of product architecture during the product development process. We classify product architectures into three categories: modular, hybrid, and integral. Existing research shows that the choice of product architecture during the new product development is a crucially strategic decision for a manufacturing firm. However, no single architecture is optimal in all cases; thus, analytic models are required to identify and discuss specific trade-offs associated with the choice of the optimal architecture under different circumstance. This chapter develops analytic models whose objectives are obtaining global performance of product through a modular/hybrid/integral architecture. Trade-offs between costs and expected benefits from different product architectures are analyzed and compared. Multifunction products and small size are used as examples to formalize the models and show the impact of the global performance characteristics. We also investigate how optimal architecture changes in response to the exogenous costs of system integrators. Some empirical implications obtained from this study show that if one considers global performance, modular architecture is an absolutely suboptimal decision and integral architecture is an all-the-time candidate for optimal architecture.KeywordsSystem IntegratorOptimal DecisionGlobal PerformanceFunctional ElementManufacturing FirmThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

This paper deals with the development of reusable FFT cores for OFDM in the light of the new features offered by modern FPGAs and synthesis tools. A pipelined and a sequential architecture have been selected as case studies to show how the requirements of various OFDM standards affect the choice of the optimal FFT architecture

Architecture selection for systems undergoing rapid technological and market change is challenging. It is desirable to select architectures that can provide cost-effective possibilities for future changes and avoid architecture lock-in. However, optimal architectures for prevailing conditions may not be changeable for future adaptation. This tension between objectives for system (product) development for both short-term and long-term competitiveness has been an enduring challenge for system architects. Here, we use time-expanded decision networks (TDNs) with time-varying costs and demands to systematically explore future architecture transition pathways and strategically identify useful designs. We demonstrate a new application for autonomous driving (AD) systems, a nascent technology, where the design and capabilities of constituent components (such as sensors, processors, and data communication links) are still evolving and significant market and regulatory uncertainties persist. In this case, we model technology costs with time-based factors to explicitly include future trends. The results show that as cost differences between architectures increase and demand for new functionality changes with time, the approach is able to identify potential transition points between architecture choices that optimize the net present value (NPV) of the system. For some of the specific scenarios analyzed in this study, the NPV with optimal architecture transitions is at least 10–20% larger as compared with fixed cases. Overall, this work presents a case for planning and partly constructing architecture transition roadmaps for new systems wherein dominant architectures have not emerged.

The multilayer perceptron has a large wide of classification and regression applications in many fields: pattern recognition, voice and classification problems. But the architecture choice has a great impact on the convergence of these networks. In the present paper we introduce a new approach to optimize the network architecture, for solving the obtained model we use the genetic algorithm and we train the network with a back-propagation algorithm. The numerical results assess the effectiveness of the theoretical results shown in this paper, and the advantages of the new modeling compared to the previous model in the literature.