The emerging issue of microplastics in marine environment: A bibliometric analysis from 2004 to 2020.

Microplastic pollution in marine environment has been a growing public concern in recent years. This article analyzed the scientific literatures related to marine microplastics through a combination of social network analysis and bibliometrics. Researches related to microplastics have grown rapidly since 2011, with approximately two-thirds of the total number of articles published in the last three years. Researchers in United States and Europe have provided tremendous support, however, the efforts and progress of Chinese researchers cannot be ignored. Moreover, the international cooperation is getting closer, and related strategies are launched continuously. The results showed that Marine Pollution Bulletin is the most active journal. Through keyword analysis, we understand the development history and current hotspots of the whole microplastics researches, including ecological risks, interrelationship between microplastics and other pollutants, and detection methodology. Finally, some suggestions and perspectives for future microplastics research are provided.

Giving waterbodies the treatment they need: A critical review of the application of constructed floating wetlands.

Water quality is declining worldwide and an increasing number of waterbodies lose their ecological function due to human population growth and climate change. Constructed floating wetlands (CFWs) are a promising ecological engineering tool for restoring waterbodies. The functionality of CFWs has been studied in-situ, in mesocosms and in the laboratory, but a systematic review of the success of in situ applications to improve ecosystem health is missing to date. This review summarises the pollutant dynamics in the presence of CFWs and quantifies removal efficiencies for major pollutants with a focus on in situ applications, including studies that have only been published in the Chinese scientific literature. We find that well designed CFWs successfully decrease pollutant concentrations and improve the health of the ecosystem, shown by lower algae biomass and more diverse fish, algae and invertebrate communities. However, simply extrapolating pollutant removal efficiencies from small-scale experiments will lead to overestimating the removal capacity of nitrogen, phosphorus and organic matter of in situ applications. We show that predicted climate change and eutrophication scenarios will likely increase the efficiency rate of CFWs, mainly due to increased growth and pollutant uptake rates at higher temperatures. However, an increase in rainfall intensity could lead to a lower efficiency of CFWs due to shorter hydraulic retention times and more pollutants being present in the particulate, not the dissolved form. Finally, we develop a framework that will assist water resource managers to design CFWs for specific management purposes. Our review clearly highlights the need of more detailed in situ studies, particularly in terms of understanding the short- and long-term ecosystem response to CFWs under different climate change scenarios.

Optimizing Autoencoders for Learning Deep Representations From Health Data.

Analyzing patients' health data using machine learning techniques can improve both patient outcomes and hospital operations. However, heterogeneous patient data (e.g., vital signs) and inefficient feature learning methods affect the implementation of machine learning-based patient data analysis. In this paper, we present a novel unsupervised deep learning-based feature learning (DFL) framework to automatically learn compact representations from patient health data for efficient clinical decision making. Real-world pneumonia patient data from the National University Hospital in Singapore are collected and analyzed to evaluate the performance of DFL. Furthermore, publicly available electroencephalogram data are extracted from the UCI Machine Learning Repository to test and support our findings. Using both data sets, we compare the performance of DFL to that of several popular feature learning methods and demonstrate its advantages.