Search

With the growing public attention on sustainable development and green ecosystems, the efficient management of fuzzy sewage treatment processes (FSTPs) has been a major concern in academia. Characterized by strong abstraction and analysis abilities, data mining technologies provide a novel perspective to solve this problem. In recent years, data-driven management for FSTP has been widely investigated, resulting in a number of typical approaches. However, almost all existing technical approaches consider FSTP a unidirectional, sequential process, ignoring the bidirectional temporality caused by backflow operations. Therefore, we propose a data-driven management mechanism for FSTP based on hybrid neural computing (IM-HNC for short). This mechanism attempts to capture the bidirectional time-series features of FSTP with the aid of a bidirectional long short-term memory model, and further introduces a convolutional neural network to construct feature spaces with a stronger expression capability. Empirically, we implement a series of experiments on three datasets under different parameter settings to test the efficiency and robustness of the proposed IM-HNC. The experimental results manifest that the IM-HNC has an average performance improvement of approximately 5% compared to the baselines. [ABSTRACT FROM AUTHOR]

With the rapid growth of the world's population and urbanization, people are increasingly seeking higher-quality medical services to improve their lives. The classification method based on deep convolutional neural networks (CNNs) is widely used in smart healthcare systems along with advancements in communication and hardware technology. Unfortunately, for conventional deep CNN algorithms, most of the regions do not participate in the convolution operation, resulting in the loss of feature information and the correlation of information between the features. To address this issue, this paper proposes a new strategy of aggregation decentralized down-sampling to prevent the loss of feature information. The regions that are not involved in the convolution operation are re-convoluted and stacked onto depth information in the forward propagation layer and the short-circuit layer, ensuring gradual convergence of the feature map and avoiding the loss of feature information. The accuracy of the proposed residual network (ResNet) system for classification tasks showed an average improvement of 2.57% compared with the conventional ResNet strategies. [ABSTRACT FROM AUTHOR]

Patients with deaths from COVID-19 often have co-morbid cardiovascular disease. Real-time cardiovascular disease monitoring based on wearable medical devices may effectively reduce COVID-19 mortality rates. However, due to technical limitations, there are three main issues. First, the traditional wireless communication technology for wearable medical devices is difficult to satisfy the real-time requirements fully. Second, current monitoring platforms lack efficient streaming data processing mechanisms to cope with the large amount of cardiovascular data generated in real time. Third, the diagnosis of the monitoring platform is usually manual, which is challenging to ensure that enough doctors online to provide a timely, efficient, and accurate diagnosis. To address these issues, this paper proposes a 5G-enabled real-time cardiovascular monitoring system for COVID-19 patients using deep learning. Firstly, we employ 5G to send and receive data from wearable medical devices. Secondly, Flink streaming data processing framework is applied to access electrocardiogram data. Finally, we use convolutional neural networks and long short-term memory networks model to obtain automatically predict the COVID-19 patient's cardiovascular health. Theoretical analysis and experimental results show that our proposal can well solve the above issues and improve the prediction accuracy of cardiovascular disease to 99.29%. [ABSTRACT FROM AUTHOR]

Since wastewater treatment processes (WTP) are generally accompanied with intense coupling and fuzziness, conventional biochemical mechanisms-based methods cannot comprehensively express the WTP due to limited computational ability. In response to the challenge caused by fuzziness, this paper proposes a hybrid control and prediction system for modeling WTP with the fuse of Activated Sludge model, Convolutional neural network and Long short-term memory neural networks (AS-CL) with knowledge and data-driven characteristics. Moreover, the activated sludge model is employed to model the wastewater treatment process based on the perspective of knowledge. Besides, the hybrid neural network that combines convolutional neural network and long short-term memory model is adopted to model the WTP from the perspective of data. Then, a multi-layer perception model is set up to realize collaborative awareness of data and knowledge. Lastly, the proposed AS-CL has been evaluated by a real-world data-set collected from a real sewage treatment plant. The results show that compared with typical existing methods, the proposal improves modeling efficiency. With the collaborative modeling scheme, influence from fuzziness on WTP can be reduced to some extent. Compared with five benchmark methods of the two evaluation indexes, the results of AS-CL show that the average performance of this method exceeds 7% of the baseline. [ABSTRACT FROM AUTHOR]

Since medical diagnosis runs through the whole journey of human health, the traditional medical diagnosis methods cannot ensure the diagnosis accurate due to the interference of multiple external factors. In response, this paper proposes a multimedia medical management method supported by the data-driven intelligent decision (MMD-DI). This method can be used to predict the survival time of cancer patients under the help of gradient boosting decision tree (GBDT) and hybrid neural network model. First, the feature factors were scanned to match the conditions by GBDT, according to the set value domain; Then the factors were inputted into the neural network. The hybrid neural network was employed to predict the survival time of cancer patients, and it was constructed by combining the convolutional neural network (CNN) and the long short-term memory (LSTM) model. Finally, the stability of the proposed MMD-DI was analyzed, and performance was compared with a series of commonly exploited baseline methods: the mean of cross-validated RMSE (Root Mean Squared Error) evaluation results is 0.183, the mean of cross-validated MAE (Mean Absolute Error) evaluation results is 0.147, the two indicators are both much lower than the commonly exploited baseline methods. A series of experiments proved that MMD-DI has excellent performance and can be used in the multimedia medical management systems. [ABSTRACT FROM AUTHOR]

Knowledge graph question answering is an important technology in intelligent human–robot interaction, which aims at automatically giving answer to human natural language question with the given knowledge graph. For the multi-relation question with higher variety and complexity, the tokens of the question have different priority for the triples selection in the reasoning steps. Most existing models take the question as a whole and ignore the priority information in it. To solve this problem, we propose question-aware memory network for multi-hop question answering, named QA2MN, to update the attention on question timely in the reasoning process. In addition, we incorporate graph context information into knowledge graph embedding model to increase the ability to represent entities and relations. We use it to initialize the QA2MN model and fine-tune it in the training process. We evaluate QA2MN on PathQuestion and WorldCup2014, two representative datasets for complex multi-hop question answering. The result demonstrates that QA2MN achieves state-of-the-art Hits@1 accuracy on the two datasets, which validates the effectiveness of our model. [ABSTRACT FROM AUTHOR]

As an important means of medical imaging, elastic imaging is an indispensable part of mobile telemedicine. Ultrasound elastography has become a research hotspot because it can accurately measure soft tissue lesions. Displacement estimation is the most important step in ultrasound elastography. At present, the phase zero search method is an accurate and fast displacement estimation method. However, when the displacement exceeds 1/4 wavelength, it is invalid. The accuracy of block matching method is not high, but it is suitable for large displacement, so it can overcome this shortcoming. It is worth noting that the quality-guided block matching method has good robustness under complex mutation conditions. It can provide prior knowledge to increase the robustness of the phase-zero search under large displacement conditions. So we propose a novel displacement estimation method for real time tissue ultrasound elastography, which combines the quality-guided block matching method and the phase-zero search method. The experimental results show that this method is more accurate, faster and robust than other displacement estimation methods. [ABSTRACT FROM AUTHOR]

This paper formulates a new paradigm, 'officialism' (Guanben Zhuyi or 官本主义), and uses it to analyse China's traditional society. The author argues that officialism is an official based doctrine and a political culture and socio-political system which takes official power and authority as its core element. Under this culture and system, relationships of official power and authority are the most important social relationships. They become the basic standard for measuring a person's social value, and they are the determining factor influencing a person's social status and social attributes. In traditional China, monarchism (Junzhu Zhuyi) (君主主义) was the highest form of officialism. Despotism is the representative characteristic of officialism. People-based doctrine (Minben Zhuyi or 民本主义) is the political ideal that is the antithesis of despotism. Democracy with the rule of law is the only way of smashing officialism and of spurring traditional political civilization to head towards modern political civilization. [ABSTRACT FROM AUTHOR]