Hasil untuk "Architecture"

M. Chen, Orhan Firat, Ankur Bapna et al.

The past year has witnessed rapid advances in sequence-to-sequence (seq2seq) modeling for Machine Translation (MT). The classic RNN-based approaches to MT were first out-performed by the convolutional seq2seq model, which was then out-performed by the more recent Transformer model. Each of these new approaches consists of a fundamental architecture accompanied by a set of modeling and training techniques that are in principle applicable to other seq2seq architectures. In this paper, we tease apart the new architectures and their accompanying techniques in two ways. First, we identify several key modeling and training techniques, and apply them to the RNN architecture, yielding a new RNMT+ model that outperforms all of the three fundamental architectures on the benchmark WMT’14 English to French and English to German tasks. Second, we analyze the properties of each fundamental seq2seq architecture and devise new hybrid architectures intended to combine their strengths. Our hybrid models obtain further improvements, outperforming the RNMT+ model on both benchmark datasets.

Udit Gupta, Mariam Elgamal, G. Hills et al.

Given the performance and efficiency optimizations realized by the computer systems and architecture community over the last decades, the dominating source of computing's carbon footprint is shifting from operational emissions to embodied emissions. These embodied emissions owe to hardware manufacturing and infrastructure-related activities. Despite the rising embodied emissions, there is a distinct lack of architectural modeling tools to quantify and optimize the end-to-end carbon footprint of computing. This work proposes ACT, an architectural carbon footprint modeling framework, to enable carbon characterization and sustainability-driven early design space exploration. Using ACT we demonstrate optimizing hardware for carbon yields distinct solutions compared to optimizing for performance and efficiency. We construct use cases, based on the three tenets of sustainable design---Reduce, Reuse, Recycle---to highlight future methods that enable strong performance and efficiency scaling in an environmentally sustainable manner.

Binghui Si, Zhenyu Ni, Jiacheng Xu et al.

Metamodeling is a promising technique for alleviating the computational burden of building energy simulation. Although various machine learning (ML) algorithms have been applied, the interactive effects of multiple factors on ML algorithm performance remain unclear. In this study, six popular ML algorithms, including ridge regression, random forest, extreme gradient boosting (XGBoost), support vector regression (SVR), k-nearest neighbor (KNN) regression and multi-layer perceptron (MLP), were analyzed for a benchmark metamodeling problem in building energy simulation under the impacts of four factors: input dimension, sample size, degree of input-output sensitivity and hyperparameter optimization (HPO) technique. The results indicated that XGBoost had high model precision and strong robustness, while KNN and SVR performed poorly on the two metrics. Increasing the sample size could mitigate the impact of the other three factors on model precision, especially for MLP. The findings will assist designers, engineers and researchers in selecting suitable ML algorithms and HPO techniques based on the dataset’s characteristics and facilitate the application of metamodeling in design optimization, sensitivity analysis and decision-making processes.

Changmei Liang, Changmei Liang, Changmei Liang et al.

The cost-benefit is a key factor when selecting an appropriate sponge city construction scheme. The research of applying intelligent technology to find cost-benefit efficient planning and construction of sponge city is urgently required. This paper established a multi-objective simulation optimization framework of sponge city construction which considered minimization of runoff control rate, pollutant control rate and life-cycle cost Non-dominated sorting genetic algorithm (NSGA-II) was successfully coupled to Storm water management model to complete the simulation-optimization process. A case study in Xining, China, was conducted to demonstrate the proposed framework. The results of this research suggested that 1) different sponge city construction schemes lead to different runoff control rates and pollutant control rates although under the same investment; 2) the runoff control rate and pollutant control rate total suspended solids decreased with the increase of the rainfall return period, while the cost of sponge city construction increased with the increase of rainfall return period. Furthermore, for T = 2-year, the sponge facility exhibited the most stable control effect on runoff and pollutants among the three different return periods (T = 2-year, 5-year, 10-year); 3) sponge city construction exhibited a “cost-benefit” efficient interval. For T = 2-year, the cost-benefit high efficiency interval of sponge city construction is calculated between 1.2 billion and 1.8 billion; for T = 5-year, the interval is between 1.2 billion and 1.8 billion, while for T = 10-year, the interval is between 1.3 billion and 2.1 billion. The above observations provide reference for reasonable and effective sponge city construction in Xining, China.

Muhammad Baginda, Tazkiyah Herdi

Yayasan Pelatihan dan Sertifikasi Perusahaan Air Minum adalah lembaga pelatihan dan sertifikasi bagi perusahaan air minum di seluruh Indonesia. Dalam proses bisnisnya, yayasan ini menyediakan jasa pelatihan dan sertifikasi bagi karyawan perusahaan air minum Indonesia yang ingin mendapatkan pelatihan dan sertifikasi. Sistem informasi yang diterapkan di yayasan masih jauh dari memadai, karena informasi yang akan dikelola masih tercatat secara manual dan belum terintegrasi, sehingga proses bisnis tidak sesuai dengan tujuan organisasi. Oleh karena itu, diperlukan perancangan arsitektur enterprise dengan menggunakan metode pengembangan arsitektur (ADM) The Open Group Architecture Framework (TOGAF) yang berfokus pada arsitektur bisnis. Hasil yang diharapkan dari penelitian ini adalah integrasi sistem informasi dan teknologi informasi ke dalam proses bisnis agar selaras dan dapat membantu proses bisnis perusahaan untuk mencapai proses bisnis yang optimal, efisien dan efektif.

Yuwei Cui, Aijun Li, Biao Duan et al.

The paper concentrates on the problem of fault-tolerant control of UAV against actuator faults from the perspective of flight control system architecture. Using backstepping control method and inverse optimization theory, the design of backstepping optimal control law was constructed. Based on the fault monitoring mechanism of vehicle management computer in the distributed flight control and control allocation system, a fault-tolerant control design method was established in the case of multiple failure modes of the actuators, which compensates the influence caused by the failures. Finally, the effectiveness of the proposed strategy was verified by numerical simulation.

Seoyoung Cho, Csaba Tóth

Service life of the pavement can be predicted with proper temperature estimation of the pavement structure. Asphalt mixture displays a different modulus upon temperature change due to its viscoelasticity. The purpose of this study is to estimate such a temperature. Methodology in here includes one with the solution of heat conduction theory of the asphalt, and the other one statistical method. Results show that there is not a significant difference between the results made by two different methods. As a result of the model performance, the error range between the observed value and the predicted value is within the range suggested by other studies, so it is judged that the performance of the model is good. However, the prediction accuracy in the month with the highest and lowest temperatures per year was low. It seems that follow-up actions on this part will be necessary in the future. This study is expected to be used in various ways for road management in Hungary, and it is expected to be a basic study for the construction of road meteorological information system.

Hui Wang, Sihang Xiao, Jianshan Wang

Previous discussions about perforated auxetic metamaterials primarily focused on the ordered systems with high degree of geometric symmetry. However, it is difficult to manufacture or retain the perfect auxetic systems in practical applications. In this paper, three types of disordered perforated auxetic systems including orientated disordered system, dimensional disordered system and complete-disordered system are explored thoroughly. The perforations of interest are oval holes which are advantageous in ensuring auxeticity, reducing stress level and improving material distribution. The designed disordered systems are fabricated by 3D printing technology and then are tested by the uniaxial tension to reveal their mechanical properties and verify the related finite element models. Thereafter, the evolution of mechanical properties of these disordered systems is investigated numerically for the varied perturbed geometric parameters such as the orientation and dimension of oval hole. The results reveal that the disordered systems still show great robustness in auxetic behavior, although the disorder in orientation and dimension exists. A high degree of symmetry in microstructure is not necessary for designing perforated auxetic systems. This provides a great convenience for the flexible and practicable design and application of perforated auxetic metamaterials.

Kengo Ando, Hiroki Iimori, Takumi Takahashi et al.

We consider the problem of uplink signal detection in scalable cell-free mMIMO (CF-mMIMO) systems subject to limited fronthaul link capacity and highly correlated channel conditions. Unlike centralized MIMO systems, in which all receive antennas are placed at a central access point (CAP), in the CF-mMIMO architecture the CAP serving a given area also uses information (<inline-formula> <tex-math notation="LaTeX">$i.e$ </tex-math></inline-formula>. channel estimates and receive signals) collected by a set of surrounding access points (APs). For such a scenario, two new robust receivers are designed, which can combat the effects of limited fronthaul capacity by leveraging knowledge of the heteroscedastic covariance of the resulting effective noise. The first receiver, which has a higher complexity but yields the best performance, is based on an expectation propagation (EP) approach, while the second employs the effective noise heteroscedastic covariance in a generalized least squares (GLS) variation of the maximum likelihood (ML) detection problem. Simulation results confirm the efficacy of both proposed receivers, which are further employed to empirically study the optimum distribution of antennas among the CAP and APs.

Vincent Bushong, Amr S. Abdelfattah, Abdullah A. Maruf et al.

Microservice architecture has become the leading design for cloud-native systems. The highly decentralized approach to software development consists of relatively independent services, which provides benefits such as faster deployment cycles, better scalability, and good separation of concerns among services. With this new architecture, one can naturally expect a broad range of advancements and simplifications over legacy systems. However, microservice system design remains challenging, as it is still difficult for engineers to understand the system module boundaries. Thus, understanding and explaining the microservice systems might not be as easy as initially thought. This study aims to classify recently published approaches and techniques to analyze microservice systems. It also looks at the evolutionary perspective of such systems and their analysis. Furthermore, the identified approaches target various challenges and goals, which this study analyzed. Thus, it provides the reader with a roadmap to the discipline, tools, techniques, and open challenges for future work. It provides a guide towards choices when aiming for analyzing cloud-native systems. The results indicate five analytical approaches commonly used in the literature, possibly in combination, towards problems classified into seven categories.

Jose E. Almanza-Medina, Benjamin Henson, Yuriy V. Zakharov

This paper investigates the use of supervised Deep Learning (DL) networks to process sonar images for underwater navigation. State-of-the-art DL techniques for micro-navigation using sequences of optical images have been adapted to work with sonar images. Specifically, the DL networks estimate the Forward-Looking Sonar (FLS) motion in three degrees of freedom corresponding to $x$ - and $y$ -translation and rotation around $z$ -axis. The state-of-the-art DL architectures and a proposed new architecture are investigated for motion estimation. They are trained using images generated by a FLS simulator. The data sets are made using pairs of consecutive images associated with labels that represent the motion of the sonar platform between images. The results show the effectiveness of using the DL architectures, which can provide millimeter accuracy for translation motion and below 0.1&#x00B0; for rotation motion between two consecutive sonar images. Examples of trajectory estimation and mosaic building using simulated and real sonar images are also presented.

Yasaldez Eder Loaiza Zuluaga, Gloria Matilde Gil Duque, Fabián Andrés David Narváez

The study shows a proposal on specific diagnosis of critical thinking skills in higher education students according to literal, inferential and critical reading levels. A literature review on Critical thinking was done to support concepts. This proposal is adapted from the variables of the information obtained. One of the conclusions shows that the career profile is a determining factor in the result and that motivation and interest are outstanding in the reading and writing processes. Teaching students to evaluate their own processes and helping teachers to include critical readings in their curricula will help to develop effective skills associated with critical thinking. The general objective of the research project is to strengthen Critical Thinking and the argumentative capacity of the students of the schools of Administration, Health Sciences, Engineering and Architecture, Social Sciences, Humanities and Theology through strategies of reading and writing. A mixed methodology was used; the researches designed and administered 2 critical reading texts, each with 10 questions and organized in 3 reading levels (literal,inferential and critical), to 158 students; then, a data analysis on how students infer, interpret and analyze text content before they start a communicative course was done. The results show that according to the professional profile, some students are more argumentative that others and that students lack Critical Thinking skills.

Archana Tiwari, Prahar Dahal, Sudeep Reg et al.

Introduction: In diagnosis of thyroid lesions, the negative pressure applied during fine needle aspiration cytology (FNAC) frequently produces bloody smears. This results in a compromise in cellular concentration and architecture which may lead to improper interpretation. Fine needle capillary sampling cytology (FNCC), on the other hand, avoids active aspiration as it depends on capillary tension to collect tissue samples in the needle bore. This study evaluated the diagnostic performance of FNAC and FNCC in thyroid lesions. Methods: A total of 120 patients were included in this study conducted over a duration of 19 months. All thyroid swellings advised for cyto-diagnosis were sampled by both fine-needle aspiration (FNAC) and non-aspiration (FNCC) techniques. The slides were assessed according to the Mair et al. scoring system. Results: In the FNCC group, 72 (60%) smears were diagnostically superior while 54 (45%) smears were diagnostically superior in the FNAC group. Blood contamination (p=0.003), cellular trauma (p=0.019), and degree of cellular degeneration (p=0.026) were less and cellular architecture (p=0.047) was preserved more in FNCC in comparison to FNAC groups. Conclusion: This study showed the superiority of FNCC for the interpretation and diagnosis of thyroid lesions. However, the combination of both FNAC and FNCC could maximize the diagnostic yield.

Xiaomei Zhong, Haibo Niu, Yongsheng Wu et al.

The M/V Marathassa oil spill occurred on 8 April 2015 in the English Bay. In the present study, the trajectory and the transport mechanism of the spilled oil have been studied by using the three-dimensional and particle-based Oil Spill Contingency and Response (OSCAR) model forced by the Finite-Volume Community Ocean Model (FVCOM). FVCOM provided the hydrodynamic variables used by the oil spill model of OSCAR. The results showed that the fraction of the oil on the water surface and on the shoreline, as well as the amount of oil recovered were affected by the time of the initial release, the overall duration of the discharge, wind and recovery actions. The hindcast study of the M/V Marathassa oil spill showed that the likely starting time for the discharge was between 14:00 and 15:00, on 8 April 2015. The release may have lasted for a relatively long time (assumed to be 22 h in this study). The results of modeling in this study were found reasonably acceptable allowing for further application in risk assessment studies in the English Bay and Vancouver Harbour. The trajectory of the spill was mainly controlled by the tidal currents, which were strongly sensitive to the local coastline and topography of First Narrows and that in the central harbour. The model results also suggested that a high-resolution model, which was able to resolve abrupt changes in the coastlines and topography, was necessary to simulate the oil spill in the harbour.

Qinrang LIU, Chongyang LIU, Jun ZHOU et al.

Concerning the issue that the convolutional neural network (CNN) accelerator design on most FPGA ends fails to effectively use the sparsity and considering both bandwidth and energy consumption,two improved CNN calculation optimization strategies based on linear systolic array architecture are proposed.Firstly,convolution is transformed into matrix multiplication to take advantage of sparsity.Secondly,in order to solve the problem of large I/O demand in traditional parallel matrix multiplier,linear systolic array is used to improve the design.Finally,a CNN acceleration comparative analysis of the advantages and disadvantages between parallel matrix multiplier and two improved linear systolic arrays is presented.Theoretical proof and analysis show that compared with the parallel matrix multiplier,the two improved linear systolic arrays make full use of sparsity,and have the advantages of less energy consumption and less I/O bandwidth occupation.

E. Weibel, D. Gómez

Jamal Abdul Naser Shokory, Jun-ichiro Giorgos Tsutsumi, Hiroyuki Yamada et al.

In Afghanistan, spring and summertime flash floods result in 54.3% of all natural disasters with an average annual economic loss of US$92.17 million between 1990 and 2014. Knowledge of climate and rainfall periodicity are urgently needed for urban and rural land use and infrastructure planning, and their flood protection. In this study, the Thornthwaite equation was used to determine the seasonal characteristics of dry and wet periods. Spectral Analysis of precipitation was carried out to search for periodicities and intraseasonal oscillations of rainfall within the Kabul River Basin. The analysis is based on data obtained from 49 ground-base stations with three parameters (precipitation, relative humidity, and temperature) for an 8-year (2006-2013) and 5-year (2009-2013) record. Five years weather maps made from JRA-55 (55 year Japan Reanalysis data) were used to recognize the moisture trajectory in each season.The results indicated that Kabul River Basin is divided into three parts with distinct climate regions as Central, Northern, and Eastern parts. The Weather maps of relative humidity with wind arrows showed the origins and pathways of air masses leading to heavy rainfall from the Arabian Sea and the Caspian Sea approaching from the west and northwest of Afghanistan in winter and spring season, corresponding to 10-day spectral peaks. In summer and autumn, influences of South Asian summer monsoon were recognized by 6-8 days of oscillation with different density in each station, where it was much significant in the Eastern part.

Christopher Heywood, Monique Arkesteijn

This paper deepens the understanding of Corporate Real Estate (CRE) alignment through a meta-study of twenty existing alignment models. A qualitative hermeneutic method interpreted the models and their articles. This holistic analysis found alignment to be more complex and pluralistic than the individual models assumed. Four dimensions operating simultaneously were evident – a multi-valent relationship, multiple alignment forms, multiple cognitive objects to align and alignment in multiple directions. Alignment theorisation had positive and negative aspects. Positive is that good science was evident and had improved over time. Negative is that model theorisation had occurred mostly in isolation and was constrained by simplifications required to make modelling tractable. The research makes a meta-theoretical contribution through a more complete theorisation of CRE alignment as a phenomenon. This addresses a disordered sense to prior theory thereby representing a major conceptual improvement. A new alignment model is not proposed; rather through developed understanding a basis is provided to examine alignment in both theorisation and practice.

Johanna Lozoya

En el presente número, Academia XXII explora en su sección temática la multidimensionalidad de la memoria como una herramienta cultural estratégica en la invención, desarrollo y control de valores socioculturales, geopolíticas del poder y discurso. Preguntas anfitrionas como qué papel estratégico juega la memoria en la reorientación de la ciencia y de la tecnología hacia la sostenibilidad medioambiental; cuáles son los límites del discurso científico social sobre la memoria cuando se construye como concepto, materia, proceso e imaginario al interior de los nichos académicos, o quién tienen el derecho de nombrar (le) en virtud de la consolidación geopolítica de valores institucionales y colectivos, han estimulado una reflexión interdisciplinaria que revela preocupaciones compartidas y propuestas estratégicas múltiples.

T. Blank