Applying Feature-Weighted Gradient Decent K-Nearest Neighbor to Select Promising Projects for Scientific Funding

Due to its outstanding ability in processing large quantity and high-dimensional data, machine learning models have been used in many cases, such as pattern recognition, classification, spam filtering, data mining and forecasting. As an outstanding machine learning algorithm, K-Nearest Neighbor (KNN) has been widely used in different situations, yet in selecting qualified applicants for winning a funding is almost new. The major problem lies in how to accurately determine the importance of attributes. In this paper, we propose a Feature-weighted Gradient Decent K-Nearest Neighbor (FGDKNN) method to classify funding applicants in to two types: approved ones or not approved ones. The FGDKNN is based on a gradient decent learning algorithm to update weight. It updatesthe weight of labels by minimizing error ratio iteratively, so that the importance of attributes can be described better. We investigate the performance of FGDKNN with Beijing Innofund. The results show that FGDKNN performs about 23%, 20%, 18%, 15% better than KNN, SVM, DT and ANN, respectively. Moreover, the FGDKNN has fast convergence time under different training scales, and has good performance under different settings.     

Effect of S on H-induced grain-boundary embrittlement in γ-Fe by first-principles calculations

The effect of interstitial impurities (H and S) on the atomic, electronic structure, and mechanical properties of the γ-Fe Σ5 (021) grain boundary (GB) was investigated via first-principles calculations. H atoms act as an intergranular embrittler in the Σ5 GB due to a reduction in the charge density between the Fe atoms connected to the grains, whereas H and S co-segregation produces more pronounced embrittlement behavior, resulting in intergranular fracture. The S-induced embrittlement plays a crucial role in the H and S segregation, due to a combination of the structural and chemical effects. The fracturing of Σ5 GB due to S and H segregation is a two-step process. The first step is the breaking of Fe–Fe bonds in the GB, followed by the breaking of the remaining Fe–S bonds in the second step, resulting in the complete separation of the two grains. Moreover, the S atom can slightly compensate for the embrittlement induced by H, because some of the Fe atoms that obtain electrons from the S atoms can provide more electrons to the H atoms, and thus, they can compensate for the electrons that must be acquired from other Fe atoms. We call this “the electrons compensating effect” and this effect is helpful in the design and alloying of steels that are resistant to H embrittlement.     

Internet use and farm households food and nutrition security nexus: The case of rural Ghana

The impact of internet use on food and nutrition security of rural households in Ghana is investigated in this study. To offset the potential challenge of selection bias, an endogenous treatment regression (ETR) technique is utilized for the analysis.. The results reveal that Internet use can improve smallholder farmers’ food and nutrition security. Internet usage has a profound positive effect on the food security of households with off-farm work and larger size of landholding. Our results suggest the intensification of efforts to enhance Internet connectivity across the nation by the government and policymakers is essential since it can go a long way to affect household welfare. The findings also highlight the importance of information and communication technologies (ICTs), the Internet, patronization to improve rural household welfare.     

A novel polysaccharide prepared from Chrysanthemum morifolium cv. Fubaiju tea and its emulsifying properties

In this study, a novel chrysanthemum polysaccharide (CP) was obtained and investigated from Chrysanthemum morifolium tea in Hubei province, China. Molecular characterisation and emulsifying properties of the CP were elucidated. The effects of different factors, including storage time, metal ions, pH value and heat treatment, were studied on the stability of emulsion. Results showed that the mean particle diameter (MPD) (d₃₂) varied from 5.54 to 0.49 μm with the polysaccharide concentration ranging from 0.5wt% to 3.0wt%. The MPD (d₃₂) of emulsions stabilised by chrysanthemum tea polysaccharide (ECP) was smaller than that of GA at 2wt%, and the MPD (d₃₂) of ECP remained stable with no stratification occurred at 25 °C for 10 days. ECP also showed good stability under acidic conditions (3.0–7.0). With increase of Na⁺ and Ca²⁺ concentration, MPD(d₃₂) and zeta potential values raised significantly, from 2.24 to 7.24 μm and from −40.52 to −35.35 mV respectively. After high-temperature heat treatment, MPD (d₃₂) and zeta potential showed no significant change, indicating ECP presented good storage stability. In conclusion, CP helped stabilised emulsions that could be used as a natural emulsifier for future application.     

Hydrogen release mechanisms of MgH2 over NiN4-embedded graphene nanosheet: First-principles calculations

In this study, first-principles calculations were performed to investigate the catalytic effect of NiN4-G on the dehydrogenation of MgH₂. Side-on MgH₂ can be adsorbed stably on the NiN4-G monolayer and is preferentially adsorbed on the NiN4 site compared with the graphene site. The hydrogen desorption process, in which MgH₂ dissociated to the Mg atom on the NiN4 site or graphene site and an H₂ molecule in the vacuum, should overcome lower barriers than pure MgH₂. This is because the corresponding Mg–H bond is weakened owing to the electron transfer between the Mg atom and the substrate. The hydrogen desorption enthalpies of the (MgH₂)₅ cluster on the NiN4 active and graphene sites are significantly smaller (0.11 eV and 1.50 eV, respectively) when H2+H2 is released from the cluster compared with those of the undoped MgH₂ cluster (2.48 eV). Therefore, the NiN4-G monolayer can provide the double effect of the NiN4 active and graphene sites on improving the dehydrogenation performance of MgH₂.     

Efficient utilization of silanol group in rice husk to promote the synthesis of Ni-phyllosilicate for CO2 methanation

It is a challenge to prepare Ni-phyllosilicate catalysts deriving from the reaction of nickel salts and biomass-based silica with high Ni contents. In order to efficiently synthesize nickel phyllosilicate, the acid-treated rice husk without calcination (P-R) was directly used as the raw material. The as-synthesized Ni-phyllosilicate (N/P-R) exhibited high nickel content of 46.7 wt% owing to utilization of the abundant silanol group in the fresh rice husk. On the contrary, the Ni content of Ni-phyllosilicate sample (N/R–S) prepared using the calcinated rice husk-based silica was only 4.8 wt% because of the loss of silanol group during calcination process. As a result, N/P-R displayed much higher catalytic activity for CO₂ methanation than N/R–S. It was found that the silanol group was in favour of forming the important intermediate Si(OH)₄, and thereby facilitated the formation of nickel phyllosilicate. In addition, the N/P-R catalyst demonstrated superior stability and excellent anti-sintering ability in the long lifetime test. In short, the proposed method in this work could not only efficiently synthesize nickel phyllosilicate with high Ni content, but also reduce energy consumption and operation steps compared to the one prepared using biomass-based silica.     

Electrocatalytic hydrogen evolution of highly dispersed Pt/NC nanoparticles derived from porphyrin MOFs under acidic and alkaline medium

Platinum (Pt)-based electrocatalysts often exhibit an excellent electrocatalytic activity for hydrogen evolution due to their best hydrogen bonding energy, but materials with high Pt content can easily cause aggregation and reduce their efficiency. In this article, Pt-based electrocatalysts with uniform dispersion and high efficiency were fabricated by a successive hydrothermal and calcination method, and the uniformly distributed Pt nanoparticle was mainly due to the anchoring effect of porphyrin ring in porphyrin MOFs, which made the minimized agglomerations as much as possible in the process of high temperature calcination. The results showed that the optimal Pt/NC-850 could reach current density of 10, 50 and 100 mA cm^?2 with only 17, 82 and 152 mV overpotentials, while the Pt/NC-850 also exhibited a superior stability and durability under acidic medium. Impressively, the Pt/NC-850 also had an excellent activity, good stability and durability under alkaline medium.