Deep Clustering: A Comprehensive Survey.

Cluster analysis plays an indispensable role in machine learning and data mining. Learning a good data representation is crucial for clustering algorithms. Recently, deep clustering (DC), which can learn clustering-friendly representations using deep neural networks (DNNs), has been broadly applied in a wide range of clustering tasks. Existing surveys for DC mainly focus on the single-view fields and the network architectures, ignoring the complex application scenarios of clustering. To address this issue, in this article, we provide a comprehensive survey for DC in views of data sources. With different data sources, we systematically distinguish the clustering methods in terms of methodology, prior knowledge, and architecture. Concretely, DC methods are introduced according to four categories, i.e., traditional single-view DC, semi-supervised DC, deep multiview clustering (MVC), and deep transfer clustering. Finally, we discuss the open challenges and potential future opportunities in different fields of DC.

The impacts of cold plasma on the taste and odor formation of dried silver carp products.

This study investigated non-thermal pretreatment (cold plasma, CP) on the flavor (taste and odor) profiles of dried fish products. CP treatment of 5 min contributed to accumulation of umami nucleotides adenosine 5'-monophosphate (AMP) from 30.96 to 40.82 µg/g and inosine 5'-monophosphate (IMP) from 2009.29 to 2132.23 µg/g, and significant reduction of bitter hypoxanthine ribonucleoside (HxR) and hypoxanthine (Hx), respectively (P < 0.05) in the dried fish products. A noticeable enhancement in sweet glycine (from 429.41 to 490.03 mg/100 g) and umami glutamic acid (from 55.68 to 67.76 mg/100 g) accompanied with the CP treatment (P < 0.05) based on taste activity value (TAV > 1). And the characteristic odor volatiles (nonanal, hexanal and 1-octen-3-ol) were strengthened 2.13-, 2.16- and 2.17- folds, respectively (P < 0.05). The results of equivalent umami concentration and Gibbs free energy calculation, combining with the correlation analysis, indicate that nucleotides and free amino acids synergically enhanced the taste improvement of dried fish products. Moderate lipids oxidation favored the formation of characteristic volatiles. The CP pretreatment offered new strategies for enhancing flavor of dried fish products.

Impacts of Natural Organic Matter Adhesion on Irreversible Membrane Fouling during Surface Water Treatment Using Ultrafiltration.

To understand impacts of organic adhesion on membrane fouling, ultrafiltration (UF) membrane fouling by dissolved natural organic matter (NOM) was investigated in the presence of background cations (Na+ and Ca2+) at typical concentrations in surface water. Moreover, NOM adhesion on the UF membrane was investigated using atomic force microscopy (AFM) with colloidal probes and a quartz crystal microbalance with dissipation monitoring (QCM-D). The results indicated that the adhesion forces at the NOM-membrane interface increased in the presence of background cations, particularly Ca2+, and that the amount of adhered NOM increased due to reduced electrostatic repulsion. However, the membrane permeability was almost not affected by background cations in the pore blocking-dominated phase but was aggravated to some extent in the cake filtration-governed phase. More importantly, the irreversible NOM fouling was not correlated with the amount of adhered NOM. The assumption for membrane autopsies is doubtful that retained or adsorbed organic materials are necessarily a primary cause of membrane fouling, particularly the irreversible fouling.