ABSTRACT
This article introduced a new deep learning framework for fault diagnosis in electrical power systems. The framework integrates the convolution neural network and different regression models to visually identify which faults have occurred in electric power systems. The approach includes three main steps: 1) data preparation;2) object detection;and 3) hyperparameter optimization. Inspired by deep learning and evolutionary computation (EC) techniques, different strategies have been proposed in each step of the process. In addition, we propose a new hyperparameters optimization model based on EC that can be used to tune parameters of our deep learning framework. In the validation of the framework's usefulness, experimental evaluation is executed using the well known and challenging VOC 2012, the COCO data sets, and the large NESTA 162-bus system. The results show that our proposed approach significantly outperforms most of the existing solutions in terms of runtime and accuracy.
ABSTRACT
BACKGROUND: Technical information regarding health-related advances is sometimes esoteric for the general public. News media, therefore, plays a key role in public health promotion via health information conveyance. In this study, we use China as a sample country and analyze the claims and frames in news coverage of health-related advances, with special focus on news coverage of the development and performance of newly developed or tested drugs. METHODS: A keyword search was performed to retrieve news articles from four representative news agencies in China. In total, 3029 news reports were retrieved, of which 128 were selected for further analysis. RESULTS: Four aspects of news coverage of drug development were identified: (1) the characteristics of new drugs covered, (2) the sources of information, (3) the accuracy of health information in newspapers, and (4) textual features of news coverage. CONCLUSIONS: Our findings reveal that guidelines should be established to facilitate more systematic news reporting on health-related advances. Additionally, literacy among the general public and professionalism in health information conveyance should be promoted to negate the "illusion of knowing" about health-related advances.
Subject(s)
Mass Media , Pharmaceutical Preparations , China , Health Promotion , Humans , Public HealthABSTRACT
BACKGROUND: The COVID-19 outbreak has tremendously impacted the world. The number of confirmed cases has continued to increase, causing damage to society and the economy worldwide. The public pays close attention to information on the pandemic and learns about the disease through various media outlets. The dissemination of comprehensive and accurate COVID-19 information that the public needs helps to educate people so they can take preventive measures. OBJECTIVE: This study aimed to examine the dissemination of COVID-19 information by analyzing the information released by the official WeChat account of the People's Daily during the pandemic. The most-read COVID-19 information in China was summarized, and the factors that influence information dissemination were studied to understand the characteristics that affect its dissemination. Moreover, this was conducted in order to identify how to effectively disseminate COVID-19 information and to provide suggestions on how to manage public opinion and information governance during a pandemic. METHODS: This was a retrospective study based on a WeChat official account. We collected all COVID-19-related information, starting with the first report about COVID-19 from the People's Daily and ending with the last piece of information about lifting the first-level emergency response in 34 Chinese provinces. A descriptive analysis was then conducted on this information, as well as on Qingbo Big Data's dissemination index. Multiple linear regression was utilized to study the factors that affected information dissemination based on various characteristics and the dissemination index. RESULTS: From January 19 to May 2, 2020, the People's Daily released 1984 pieces of information; 1621 were related to COVID-19, which mainly included headline news items, items with emotional content, and issues related to the pandemic's development. By analyzing the dissemination index, seven information dissemination peaks were discerned. Among the three dimensions of COVID-19 information-media salience, content, and format-eight factors affected the spread of COVID-19 information. CONCLUSIONS: Different types of pandemic-related information have varying dissemination power. To effectively disseminate information and prevent the spread of COVID-19, we should identify the factors that affect this dissemination. We should then disseminate the types of information the public is most concerned about, use information to educate people to improve their health literacy, and improve public opinion and information governance.