Suchergebnisse

The 12th Workshop on Synthesis And System Integration of Mixed Information technologies (SASIMI2004), Oct. 18-19, 2004, Kanazawa, Japan, pp.422-429. ; This publication is a work of the U.S. Government as defined in Title 17, United States Code, Section 101. As such, it is in the public domain, and under the provisions of Title 17, United States Code, Section 105, may not be copyrighted. ; The availability of large, inexpensive memory has made it possible to realize numerical functions, such as the reciprocal, square root, and trigonometric functions, using a look-up table. This is much faster than by software. However, a naive look-up method requires unreasonably large memory. In this paper, we show the use of a look-up table (LUT) cascade to realize a piecewise linear approximation to the given function. Our approach yields memory of reasonable size and significant accuracy.

IntroductionThe cascade of HIV care is one of the main tools to assess the individual and public health benefits of antiretroviral therapy (ART) and identify barriers of treatment as prevention (TasP) concept realization. We aimed to characterize the changes in engagement of HIV‐positive persons in care in Russia during three years (2011–2013).MethodsWe defined seven steps in the cascade of care framework: HIV infected (estimation data), HIV diagnosed, linked to HIV care, retained in HIV care, need ART, on ART and viral suppressed (VL < 1000 copies/mL during 12 month ART). Information was extracted from the Federal AIDS Centre database and from the national monitoring forms of Rospotrebnadzor from the beginning of 2011 to 31 December 2013.ResultsNearly 668,032 HIV‐diagnosed Russian residents were alive by the end of 2013, which consisted 49% of the estimated 1,363,330 people living with HIV. Among the alive HIV‐diagnosed patients, 516,403 (77%) were linked to care and 481,783 (72%) were retained. Of 163,822 (25% of HIV diagnosed) patients who were eligible for ART, 156,858 (96%) were on treatment while 127,054 (81%) had viral suppression. However, only 19% of HIV‐diagnosed patients achieved viral suppression which is necessary to prevent viral transmission. We noted substantial improvements over time in the proportion of individuals on ART. The proportion of patients who received ART increased from 24% in 2011 to 34% in 2013. The most significant leakages of patients during three years were on steps: "HIV infected → HIV diagnosed" (loss −55% in 2011, −53% in 2012, and −51% in 2013), "HIV diagnosed → Linked to care" (−23% yearly) and "Retained in care → Need ART" (−76%, −70%, and −66%, respectively).ConclusionThe stages of HIV diagnosis and estimation of ART eligibility were the most vulnerable to leakage. Encouraging HIV testing and earlier ART initiation are needed to maximize the effects of TasP interventions and to contain the spread of HIV in Russia.

Modeling online discourse dynamics is a core activity in understanding the spread of information, both offline and online, and emergent online behavior. There is currently a disconnect between the practitioners of online social media analysis -- usually social, political and communication scientists -- and the accessibility to tools capable of examining online discussions of users. Here we present evently, a tool for modeling online reshare cascades, and particularly retweet cascades, using self-exciting processes. It provides a comprehensive set of functionalities for processing raw data from Twitter public APIs, modeling the temporal dynamics of processed retweet cascades and characterizing online users with a wide range of diffusion measures. This tool is designed for researchers with a wide range of computer expertise, and it includes tutorials and detailed documentation. We illustrate the usage of evently with an end-to-end analysis of online user behavior on a topical dataset relating to COVID-19. We show that, by characterizing users solely based on how their content spreads online, we can disentangle influential users and online bots.

ABSTRACT:Future projections of coastal erosion, which are one of the most demanded climate services in coastal areas, are mainly developed using top-down approaches. These approaches consist of undertaking a sequence of steps that include selecting emission or concentration scenarios and climate models, correcting models bias, applying downscaling methods, and implementing coastal erosion models. The information involved in this modelling chain cascades across steps, and so does related uncertainty, which accumulates in the results. Here, we develop long-term multi-ensemble probabilistic coastal erosion projections following the steps of the top-down approach, factorise, decompose and visualise the uncertainty cascade using real data and analyse the contribution of the uncertainty sources (knowledge-based and intrinsic) to the total uncertainty. We find a multi-modal response in long-term erosion estimates and demonstrate that not sampling internal climate variability?s uncertainty sufficiently could lead to a truncated outcomes range, affecting decision-making. Additionally, the noise arising from internal variability (rare outcomes) appears to be an important part of the full range of results, as it turns out that the most extreme shoreline retreat events occur for the simulated chronologies of climate forcing conditions. We conclude that, to capture the full uncertainty, all sources need to be properly sampled considering the climate-related forcing variables involved, the degree of anthropogenic impact and time horizon targeted. ; AT acknowledges the financial support from the FENIX Project by the Government of Cantabria. This research was also funded by the Spanish Government via the grant RISKCOADAPT (BIA2017-89401-R).

A theory is said to be fully absorbable whenever its own acceptance by all of the individuals belonging to a certain population does not question its predictive validity. This accounts for strategic equilibria and can be related to the logic underlying convergence of behaviour and intentional herding in sequential games. This paper discusses the absorbability of informational cascades' theory by bounded rational decision-makers and analyses whether providing individuals with theoretic information on informational cascades affects overall probability of herding phenomena to occur as well as whether an incorrect cascade can be reversed because of bounded rational adapting of the theory's prescriptive.

People memory system plays a key role in the decision‐making process (DMP). In order to examine its influence on decision‐making, an individual's memory‐based decision‐making system is developed, and then, we integrate the multi‐agent decision systems and construct a sequential risky decision model on the basis of an informational cascade. This study aims at exploring the public's decisions on whether or not to take protective actions under risk. The findings indicate that people with different strength of ties to friends and relatives make huge differences on decision‐making. In the group with weak ties, the agent's total size of decision information and the level of risk perception are reducing in the sequential decision process. We further prove that people with weak ties are more prone to take no protective action, and the probability is also decreasing with the decision turn. The sequence of people with strong ties makes decisions dynamically with the intensity of released information. Finally, the influences of forgetting rate and memory capacity on people's decision‐making are examined. The model provides a new line of thought about building a multi‐agent system of DMP, which is also very helpful for the design of an information management system during emergencies. Copyright © 2014 John Wiley & Sons, Ltd.

Automated detection of the content of images remains a challenging problem in artificial intelligence. Hence, continuous manual monitoring of restricted development zones is critical to maintaining territorial integrity and national security. In this regard, local governments of the Republic of Korea conduct four periodic inspections per year to preserve national territories from illegal encroachments and unauthorized developments in restricted zones. The considerable expense makes responding to illegal developments difficult for local governments. To address this challenge, we propose a deep-learning-based Cascade Mask region-based convolutional neural network (R-CNN) algorithm designed to perform automated detection of greenhouses in aerial photographs for efficient and continuous monitoring of restricted development zones in the Republic of Korea. Our proposed model is regional-based because it was optimized for the Republic of Korea via transfer learning and hyperparameter tuning, which improved the efficiency of the automated detection of greenhouse facilities. The experimental results demonstrated that the mAP value of the proposed Cascade Mask R-CNN model was 83.6, which was 12.83 higher than baseline mask R-CNN, and 0.9 higher than Mask R-CNN with hyperparameter tuning and transfer learning considered. Similarly, the F1-score of the proposed Cascade Mask R-CNN model was 62.07, which outperformed those of the baseline mask R-CNN and the Mask R-CNN with hyperparameter tuning and transfer learning considered (i.e., the F1-score 52.33 and 59.13, respectively). The proposed improved Cascade Mask R-CNN model is expected to facilitate efficient and continuous monitoring of restricted development zones through routine screening procedures. Moreover, this work provides a baseline for developing an integrated management system for national-scale land-use planning and development infrastructure by synergizing geographical information systems, remote sensing, and deep learning models.

Automated detection of the content of images remains a challenging problem in artificial intelligence. Hence, continuous manual monitoring of restricted development zones is critical to maintaining territorial integrity and national security. In this regard, local governments of the Republic of Korea conduct four periodic inspections per year to preserve national territories from illegal encroachments and unauthorized developments in restricted zones. The considerable expense makes responding to illegal developments difficult for local governments. To address this challenge, we propose a deep-learning-based Cascade Mask region-based convolutional neural network (R-CNN) algorithm designed to perform automated detection of greenhouses in aerial photographs for efficient and continuous monitoring of restricted development zones in the Republic of Korea. Our proposed model is regional-based because it was optimized for the Republic of Korea via transfer learning and hyperparameter tuning, which improved the efficiency of the automated detection of greenhouse facilities. The experimental results demonstrated that the mAP value of the proposed Cascade Mask R-CNN model was 83.6, which was 12.83 higher than baseline mask R-CNN, and 0.9 higher than Mask R-CNN with hyperparameter tuning and transfer learning considered. Similarly, the F1-score of the proposed Cascade Mask R-CNN model was 62.07, which outperformed those of the baseline mask R-CNN and the Mask R-CNN with hyperparameter tuning and transfer learning considered (i.e., the F1-score 52.33 and 59.13, respectively). The proposed improved Cascade Mask R-CNN model is expected to facilitate efficient and continuous monitoring of restricted development zones through routine screening procedures. Moreover, this work provides a baseline for developing an integrated management system for national-scale land-use planning and development infrastructure by synergizing geographical information systems, remote sensing, and deep learning models.

The demand for the sensor-based detection of camouflage objects widely exists in biological research, remote sensing, and military applications. However, the performance of traditional object detection algorithms is limited, as they are incapable of extracting informative parts from low signal-to-noise ratio features. To address this problem, we propose Camouflaged Object Detection with Cascade and Feedback Fusion (CODCEF), a deep learning framework based on an RGB optical sensor that leverages a cascaded structure with Feedback Partial Decoders (FPD) instead of a traditional encoder–decoder structure. Through a selective fusion strategy and feedback loop, FPD reduces the loss of information and the interference of noises in the process of feature interweaving. Furthermore, we introduce Pixel Perception Fusion (PPF) loss, which aims to pay more attention to local pixels that might become the edges of an object. Experimental results on an edge device show that CODCEF achieved competitive results compared with 10 state-of-the-art methods.

The demand for the sensor-based detection of camouflage objects widely exists in biological research, remote sensing, and military applications. However, the performance of traditional object detection algorithms is limited, as they are incapable of extracting informative parts from low signal-to-noise ratio features. To address this problem, we propose Camouflaged Object Detection with Cascade and Feedback Fusion (CODCEF), a deep learning framework based on an RGB optical sensor that leverages a cascaded structure with Feedback Partial Decoders (FPD) instead of a traditional encoder–decoder structure. Through a selective fusion strategy and feedback loop, FPD reduces the loss of information and the interference of noises in the process of feature interweaving. Furthermore, we introduce Pixel Perception Fusion (PPF) loss, which aims to pay more attention to local pixels that might become the edges of an object. Experimental results on an edge device show that CODCEF achieved competitive results compared with 10 state-of-the-art methods.