Search

This paper investigates the correlation between the persistent impact of the epidemic on investment risk in the post-epidemic period and the response measures. A high correlation is found between market performance and policies introduced due to the ongoing impact of the epidemic. Further validation is provided by using the CSI 300 as a reference portfolio and by selecting ten stocks from the CSI 300 in the capital markets services sector to calculate the correlation coefficient between the latter and the former. It is found that placing assets in capital market services companies leads to lower investment risk, and the correlation between investment risk and the ongoing impact of the epidemic is moderately low. However, placing assets in capital markets services firms is not a better choice currently if investors want to obtain a higher return on their assets.

Business valuations are often used today by stakeholders to estimate the economic value of the benefits a business generates. In this case, financial market participants use a variety of valuation methods to determine the price they are willing to pay or accept to achieve a stock sale. Therefore, this paper conducts financial statement analysis and common valuation methods on the basis of Airbus annual report, and then evaluates some other factors that affect valuation. The study found that Airbus got rid of high leverage and implemented sustainable development in recent years. The stock price based on common valuation methods is lower than the market value. However, this paper believes that due to the current situation of Airbus and the capital market, the current market value is reasonable.

© 2022 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting /republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works Parameter estimation is vital for modeling and control of fuel cell systems. However, the nonlinear parameterization is an intrinsic characteristic in the fuel cell models such that classical parameter estimation schemes developed for linearly parameterized systems cannot be applied. In this paper, an alternative framework of adaptive parameter estimation is designed to address the real-time parameter estimation for fuel cell systems. The parameter estimation can be divided into two cascaded components. First, the dynamics with the unknown parameters are estimated by a new unknown system dynamics estimator (USDE). Inspired by an invariant manifold, this USDE is designed by applying simple filter operations such that the information of the state derivative is not required. Secondly, an adaptive law driven by the function approximation error is proposed for recovering unknown model parameters. Exponential convergence of the estimated parameters to the true values can be proved under the monotonicity condition. Finally, experimental results on a practical proton exchange membrane fuel cell system are given to verify the effectiveness of the proposed schemes.

Book value is the value presented on the balance sheet. In contrast to book value, which reflects the past value, market value is based on the current value of the enterprise and includes all price influencing factors in the market. An accurate valuation of an enterprise is critical to making correct investment decisions. The gap between the book value and market value of an enterprise can lead to deviations in enterprise value valuation, which can affect the judgment of financial statement users. However, the current accounting standards cannot meet such decision-making needs of information users and need to be improved in the future reform. The purpose of this paper is to analyze the reasons for the gap at the empirical level and to assess the implications for the reform of accounting standards.

ObjectiveThe oblique lateral interbody fusion (OLIF) technique is a promising interbody fusion technique. This study summarizes the technical aspects of OLIF as a salvage surgery and the preliminary outcomes of a series of cases.Patients and methodsA retrospective review of patients with leg or back pain induced by pseudoarthrosis or adjacent segment disease after posterior lumbar interbody fusion/transforaminal lumbar interbody fusion was done. These patients underwent salvage OLIF surgeries in our institution from January 2021 to March 2022. Variables such as the demographic, clinical, surgical, and radiological characteristics of the enrolled patients were recorded and analyzed.ResultsEight patients (five females and three males; mean age 69.1 ± 5.7 years, range 63–80 years) were enrolled in this study. The mean operative time was 286.25 min (range: 230–440 min), and the estimated blood loss was 90 ml (range: 50–150 ml). Only one of the eight patients experienced a complication of lower limb motor weakness, which disappeared within 5 days after surgery. The latest data showed that the mean intervertebral space height increased from 8.36 mm preoperatively to 12.70 mm and the mean segmentary lordosis increased from 8.92° preoperatively to 15.05°. Bone fusion was achieved in all but one patient, who was followed up for only 3 months. The JOA scores Japanese Orthopaedic Association (JOA) Scores for low back pain of all patients significantly improved at the final follow-up.ConclusionOLIF provides a safe and effective salvage strategy for patients with failed posterior intervertebral fusion surgery. Patients effectively recovered intervertebral and foraminal height with no additional posterior direct decompression.

Large-scale classification poses great challenges to storage and computation. There are two major solutions to address the problem: data compression and quantization. In the paper, we study the method of first reducing data dimension by random projection and then quantizing the projections to ternary and binary codes, which has been widely applied in practice. Often, the extreme quantization would degrade the accuracy of classification due to high quantization errors. Interestingly, however, we observe that the quantization could result in performance improvement, rather than degradation, if the data for quantization are preprocessed by sparse transform. Also, the quantization gain could be obtained with the random projections of the data, if both the data and random projection matrices are sparse enough, such that the resulting projections remain sparse. The intriguing performance is verified and analyzed with extensive experiments.

Carbon nanotubes (CNTs) are ideal candidates for beyond-silicon nano-electronics because of their high mobility and low-cost processing. Recently, assembled massively aligned CNTs have emerged as an important platform for semiconductor electronics. However, realizing sophisticated complementary nano-electronics has been challenging due to the p-type nature of carbon nanotubes in air. Fabrication of n-type behavior field effect transistors (FETs) based on assembled aligned CNT arrays is needed for advanced CNT electronics. Here in this paper, we report a scalable process to make n-type behavior FETs based on assembled aligned CNT arrays. Air-stable and high-performance n-type behavior CNT FETs are achieved with high yield by combining the atomic layer deposition dielectric and metal contact engineering. We also systematically studied the contribution of metal contacts and atomic layer deposition passivation in determining the transistor polarity. Based on these experimental results, we report the successful demonstration of complementary metal-oxide-semiconductor inverters with good performance, which paves the way for realizing the promising future of carbon nanotube nano-electronics.

Apple xyloglucan (XyG) structural features (molecular weight and building motifs) were studied in relation with the hemicellulose interaction with cellulose nanofiber (CNF) through adsorption isotherm, high performance size exclusion chromatography and high-performance anion exchange chromatography analyses. Up to 75.6 μg of apple XyG adsorbed/mg of CNF. XyG adsorption depended on both Mw and XyG/CNF ratio. The adsorption capacity increased with Mw irrespectively of XyG/CNF ratio. At XyG/CNF ratio 0.05, the adsorption of high Mw populations (2.0 × 10

Orientation-controlled growth of two-dimensional (2D) transition metal dichalcogenides (TMDCs) may enable many new electronic and optical applications. However, previous studies reporting aligned growth of WSe2 usually yielded very small domain sizes. Herein, we introduced gold vapor into the chemical vapor deposition (CVD) process as a catalyst to assist the growth of WSe2 and successfully achieved highly aligned monolayer WSe2 triangular flakes grown on c-plane sapphire with large domain sizes (130 µm) and fast growth rate (4.3 µm·s−1). When the aligned WSe2 domains merged together, a continuous monolayer WSe2 was formed with good uniformity. After transferring to Si/SiO2 substrates, field effect transistors were fabricated on the continuous monolayer WSe2, and an average mobility of 12 cm2·V−1·s−1 was achieved, demonstrating the good quality of the material. This report paves the way to study the effect of catalytic metal vapor in the CVD process of TMDCs and contributes a novel approach to realize the growth of aligned TMDC flakes.

In the work, we study the exemplar-based classification method, which categorizes novel objects by comparing their similarity to the exemplars previously stored by class. The exemplar storage and similarity calculation will pose great challenges for large-scale classification. To reduce data complexity, we propose to first reduce data dimension by random projection and then quantize the projections to binary or ternary codes. Intuitively, the extreme quantization will degrade the accuracy of classification due to significant information loss. Interestingly, however, we observe that the quantization could provide comparable or even superior accuracy, if the data to be quantized are the sparse features generated with common filters. Furthermore, this quantization property could be preserved in the random projections of sparse features, if both the features and random projection matrices are sufficiently sparse. By conducting extensive experiments, we validate and analyze this intriguing property.

Long noncoding RNAs (lncRNAs) have been demonstrated to play key roles in various biological processes. However, the contributions of lncRNAs to Seneca Valley virus (SVV) infection and host defense remain largely unknown. In this study, differentially expressed lncRNAs and mRNAs in SVV-infected PK15 cells were detected by genome-wide analysis. A total of 14,127 lncRNAs and 63,562 mRNAs were identified, and 1,780 lncRNAs were differentially expressed. The functional prediction of SVV-induced lncRNAs showed high associations with biological regulation and many immunity-related signaling pathways, including the B-cell receptor pathway, RIG-I-like receptor signaling pathway, and NF-kappa B (NF-κB) signaling pathway. We next screened lncRNAs and target genes related to immune response pathways and further demonstrated their differential expression in SVV-infected PK15 cells. Our study investigated the function of lncRNAs involved in SVV infection and provided new insight into the pathogenic mechanisms of SVV.

Within the apple pomace biorefinery cascade processing framework aiming at adding value to an agroindustrial waste, after pectin recovery, this study focused on hemicellulose. The structure of the major apple hemicellulose, xyloglucan (XyG), was assessed as a prerequisite to potential developments in industrial applications. DMSO-LiCl and 4 M KOH soluble hemicelluloses from pectin-extracted apple pomace were purified by anion exchange chromatography. XyG structure was assessed by coupling xyloglucanase and endo-β-1,4-glucanase digestions to HPAEC and MALDI-TOF MS analyses. 71.9% of pomaces hemicellulose were recovered with starch. DMSO-LiCl and 4 M KOH soluble XyG exhibited Mw of 19 and 140 kDa, respectively. Besides the XXXG, XLXG, XXLG, XXFG, XLFG and XLLG structures, novel oligosaccharides with degree of polymerization of 6-10 were observed after xyloglucanase digestion. Cellobiose and cellotriose were revealed randomly distributed in XyG backbone and were more present in DMSO-LiCl soluble XyG. Residual pomace remains a potential source of other materials.

Extracellular vesicles (EVs) are nanoscale membrane vesicles which contain a wide variety of cargo such as proteins, miRNAs and lipids. A growing body of evidence suggests that EVs are promising biomarkers for disease diagnosis and therapeutic strategies. Although the excellent clinical value, their use in personalized healthcare practice is not yet feasible due to their highly heterogeneous nature. Taking the difficulty of isolation and the small size of EVs into account, the characterization of EVs at a single-particle level is both imperative and challenging. In a bid to address this critical point, more research has been directed into a microfluidic platform because of its inherent advantages in sensitivity, specificity and throughput. This review discusses the biogenesis and heterogeneity of EVs and takes a broad view of state-of-the-art advances in microfluidics-based EV research, including not only EV separation, but also the single EV characterization of biophysical detection and biochemical analysis. To highlight the advantages of microfluidic techniques, conventional technologies are included for comparison. The current status of artificial intelligence (AI) for single EV characterization is then presented. Furthermore, the challenges and prospects of microfluidics and its combination with AI applications in single EV characterization are also discussed. In the foreseeable future, recent breakthroughs in microfluidic platforms are expected to pave the way for single EV analysis and improve applications for precision medicine. This article is protected by copyright. All rights reserved.

Developing convenient and accurate SARS-CoV-2 antigen test and serology test is crucial in curbing the global COVID-19 pandemic. In this work, we report an improved indium oxide (InSupplementary material (materials and methods for device fabrication, functionalization of In

Recently, the method that learns networks layer by layer has attracted increasing interest for its ease of analysis. For the method, the main challenge lies in deriving an optimization target for each layer by inversely propagating the global target of the network. The target propagation is an ill-posed problem, due to involving the inversion of nonlinear activations from low-dimensional to high-dimensional spaces. To address the problem, the existing solution is to learn an auxiliary network to specially propagate the target. However, the network lacks stability, and moreover, it leads to higher complexity for network learning. In the letter, we show that target propagation could be achieved by modeling the network's each layer with compressed sensing, without the need of auxiliary networks. Experiments show that the proposed method could achieve better performance than the auxiliary network-based method.

International audience; Structural elucidation of plant cell wall xyloglucan through the analysis of enzymatically produced fragmentsrequires detailed knowledge of enzymes hydrolytic mechanism. In this note, the mode of action and cleavage siteof commercial recombinant xyloglucanases (GH74, Paenibacillus sp.) was studied on native and fluorescent-tagged tamarind xyloglucan. In complement to information provided by the manufacturer, GH74 hydrolysiswas shown dual endo/exo- and exo-processive with low affinity towards labelled reducing-ends. GH74 accom-modated X/G in its subsite --1 and X/L in its subsite +1. Moreover, hydrolysis kinetic indicated a GH74 activityinhibition by excess products. These results will help for application of this enzyme in xyloglucans structuralanalysis or for processing cell walls.

Adjacent vertebral compression fracture (AVCF) is the primary factor affecting satisfaction after PKP surgery. In addition to osteoporosis, certain structural characteristics of the vertebral body itself also increase their risk. The purpose of this study was to explore the impact of biomechanical changes on AVCF after balancing other factors.As a routine treatment of OVCF, the postoperative refracture of PKP is a serious problem. The aim of our study was to explore the impact of lumbar biomechanical changes on the risk of AVCF. A propensity score matching was performed to balance the interference of osteoporosis, which is the primary risk factor of AVCF.A retrospective, single-center case-control study was performed. From September 2013 to March 2020, 1752 patients were enrolled, and AVCF was assessed in 80 of these patients. A propensity score matching (PSM) analysis was performed, and 5 potential confounding factors were matched (age, BMI, number of fractured vertebral bodies, fracture region, and HDL). The preoperative and postoperative radiological factors were measured in the matched cohort of 48 pairs. A conditional logistic regression analysis to adjust the comparative risks.The preoperative wedge angle and its postoperative recovery of the AVCF group were significantly higher than that of the non-AVCF group. The local kyphosis of the fractured vertebral body between the two groups was similar, but the recovery in the AVCF group was slightly higher than that in the non-AVCF group. The preoperative and postoperative relative anterior height (RAH) of the fractured vertebral body was familiar in two groups, so was the recovery of RAH. The preoperative spino-sacral angle (SSA) was significantly higher in the AVCF group than in the non-AVCF group. The preoperative wedge angle was identified as the only significant risk factor for AVCF in the multivariate analysis.In conclusion, a larger preoperative fracture vertebral wedge angle is a risk factor for AVCF. For such high-risk patients, surgeons should be cautious about surgical decisions. The postoperative active measures and tailored surveillance should be attached to great importance as well.

Nonvolatile, flexible artificial synapses that can be used for brain-inspired computing are highly desirable for emerging applications such as human-machine interfaces, soft robotics, medical implants, and biological studies. Printed devices based on organic materials are very promising for these applications due to their sensitivity to ion injection, intrinsic printability, biocompatibility, and great potential for flexible/stretchable electronics. Herein, we report the experimental realization of a nonvolatile artificial synapse using organic polymers in a scalable fabrication process. The three-terminal electrochemical neuromorphic device successfully emulates the key features of biological synapses: long-term potentiation/depression, spike timing-dependent plasticity learning rule, paired-pulse facilitation, and ultralow energy consumption. The artificial synapse network exhibits an excellent endurance against bending tests and enables a direct emulation of logic gates, which shows the feasibility of using them in futuristic hierarchical neural networks. Based on our demonstration of 100 distinct, nonvolatile conductance states, we achieved a high accuracy in pattern recognition and face classification neural network simulations.

Severe lithium (Li) dendrite growth during the plating/stripping process, which results in low Coulombic efficiency and safety issues, strongly hinders the practical applications of Li metal anodes. Herein, we propose a novel design of three-dimensional (3D) interconnected graphene (IG) framework synthesized with the help of nickel (Ni) microspheres for stable Li metal anodes. The as-prepared IG framework consists of multiple stacks of two-dimensional (2D) graphene layers and plenty of hollow graphene microspheres in between, and thus provides protective layers on the top to suppress lithium dendrites, sufficient surface area to reduce the effective current density, as well as ion channels for fast Li transport, which is confirmed by post-cycle morphology characterization. When Li foil was used as the counter electrode for Li deposition, the assembled coin cell maintained an average Coulombic efficiency of more than 97.5% for 100 cycles at current density of 1 mA cm−2 with a Li loading of 1 mAh cm−2. Furthermore, we achieved stable cycling for more than 300 h at a current density of 1 mA cm−2 with a Li loading of 2 mAh cm−2 when assembled as symmetric cells. This strategy of vertically stacking 2D materials provides a novel approach towards dendrite-free Li metal anodes for the next-generation energy storage systems.

Cockpit Voice Recorder (CVR) records all the sound signals related to the accident in the cockpit before the crash. Accurate analysis of the cockpit voice signal is the basis for exploring the real reason behind the flight accident, which often starts from the separation of cockpit mixed signals. In this paper, a Fast Independent Component Analysis (FastICA) method for mixed signal separation is proposed. The basic idea of the method, three classical assumptions and the independence criteria of separation results are presented. The FastICA algorithm is analyzed theoretically, and the blind separation experiment of cockpit mixed signals is realized in this paper. The simulation experiment result shows that although FastICA algorithm is complex in theory, the realize of this algorithm is concise and clear, which can quickly and effectively separate these signals, and provide research basis for accurate analysis of the causes of aircraft accidents in the later stage.

Recently, ithas been observed that {0,±1}‐ternary codes, which are simply generated from deep features by hard thresholding, tend to outperform {−1,1}‐binary codes in image retrieval. To obtain better ternary codes, the authors for the first time propose to jointly learn the features with the codes by appending a smoothed function to the networks. During training, the function could evolve into a non‐smoothed ternary function by a continuation method, and then generate ternary codes. The method circumvents the difficulty of directly training discrete functions and reduces the quantization errors of ternary codes. Experiments show that the proposed joint learning indeed could produce better ternary codes. For the first time, the authors propose to generate ternary hash codes by jointly learning the codes with deep features via a continuation method. Experiments show that the proposed method outperforms existing methods.

Summary Flexible sensors are essential for advancing implantable and wearable bioelectronics toward monitoring chemical signals within and on the body. Developing biosensors for monitoring multiple neurotransmitters in real time represents a key in vivo application that will increase understanding of information encoded in brain neurochemical fluxes. Here, arrays of devices having multiple In2O3 nanoribbon field-effect transistors (FETs) were fabricated on 1.4-μm-thick polyethylene terephthalate (PET) substrates using shadow mask patterning techniques. Thin PET-FET devices withstood crumpling and bending such that stable transistor performance with high mobility was maintained over >100 bending cycles. Real-time detection of the small-molecule neurotransmitters serotonin and dopamine was achieved by immobilizing recently identified high-affinity nucleic-acid aptamers on individual In2O3 nanoribbon devices. Limits of detection were 10 fM for serotonin and dopamine with detection ranges spanning eight orders of magnitude. Simultaneous sensing of temperature, pH, serotonin, and dopamine enabled integration of physiological and neurochemical data from individual bioelectronic devices., Graphical Abstract, Highlights • We fabricated flexible In2O3 nanoribbon transistors using cleanroom-free processes • Flexible In2O3 transistors withstood crumpling and bending with stable performance • Flexible aptamer biosensors detect neurotransmitters in real time • Multiplexed sensors monitor temperature, pH, serotonin, and dopamine simultaneously, Nanoscience; Bioelectronics; Nanotechnology Fabrication

With the progress of intelligent and digital healthcare, wearable sensors are attracting considerable attention due to their portable and real-time monitoring capabilities. Among them, wearable gas sensors, which can detect both gas markers from the human body and hazardous gas from the environment, are particularly gaining tremendous interest. To ensure the gas sensors can be worn and carried easily, most of them were fabricated on flexible substrates. However, some traditional fabrication techniques of gas sensors such as lithography and chemical vapor deposited, are incompatible with most flexible substrates due to the flexible substrates cannot endure the harsh fabricated conditions, for instance, high temperature. Therefore, fabrication techniques for wearable gas sensors are extremely limited, thus a summary of which is necessary. Here, recent advances in the fabrication techniques of wearable gas sensors are presented. Fabrication techniques included coating techniques, printing techniques, spinning techniques, and transferring techniques are discussed in detail, respectively.

Quasi-two-dimensional (2D) β-Ga2O3 is a rediscovered metal-oxide semiconductor with an ultra-wide bandgap of 4.6–4.9 eV. It has been reported to be a promising material for next-generation power and radio frequency electronics. Field effect transistors (FETs) that can switch at high voltage are key components in power and radio frequency devices, and reliable Ohmic contacts are essential for high FET performance. However, obtaining low contact resistance on β-Ga2O3 FETs is difficult since reactions between β-Ga2O3 and metal contacts are not fully understood. Herein, we experimentally demonstrate the importance of reactions at the metal/β-Ga2O3 interface and the corresponding effects of these reactions on FET performance. When Ti is employed as the metal contact, annealing of β-Ga2O3 FETs in argon can effectively transform Schottky contacts into Ohmic contacts and permit a large drain current density of ~ 3.1 mA/μm. The contact resistance (Rcontact) between the Ti electrodes and β-Ga2O3 decreased from ~ 430 to ~ 0.387 Ω·mm after annealing. X-ray photoelectron spectroscopy (XPS) confirmed the formation of oxygen vacancies at the Ti/β-Ga2O3 interface after annealing, which is believed to cause the improved FET performance. The results of this study pave the way for greater application of β-Ga2O3 in electronics.

International audience; To establish a “green” biorefinery extraction of apple pomace pectin, a sequential pretreatment with three natural deep eutectic solvents (NADES, choline chloride (CC): glycerol (G); CC: lactic acid (LA); potassium carbonate (K): G) was used prior to hot water extraction. A synergistic effect of CC:G and CC:LA pretreatments was observed and led to the highest recovery of pectin. The sequential NADES/water extraction process also provided a mean to tailor pectin main structure. It was explained as resulting from ion exchange and individual NADES components effects. The 13C solid state NMR T1ρH and THH parameters indicated a reorganization of cellulose in the residues following extraction of pectin, notably after alkaline K:G pretreatment/water extraction. Hence, sequential NADES pretreatments/water extraction represents a “green” alternative to mild mineral acid to extract pectin and to tailor its main structures, while the residual pomace can be further sources of valuable compounds and polymers.

Distributed Denial of Service (DDOS) attack is one of the most common network attacks. DDoS attacks are becoming more and more diverse, which makes it difficult for some DDoS attack detection methods based on single network flow characteristics to detect various types of DDoS attacks, while the detection methods of multi-feature DDoS attacks have a certain lag due to the complexity of the algorithm. Therefore, it is necessary and urgent to monitor the trend of traffic change and identify DDoS attacks timely and accurately. In this paper, a method of DDoS attack detection based on deep belief network feature extraction and LSTM model is proposed. This method uses deep belief network to extract the features of IP packets, and identifies DDoS attacks based on LSTM model. This scheme is suitable for DDoS attack detection technology. The model can accurately predict the trend of normal network traffic, identify the anomalies caused by DDoS attacks, and apply to solve more detection methods about DDoS attacks in the future.

In order to build a high-quality team of teachers in vocational higher education institutions, and investigate the current situation of teachers of common required courses in vocational higher education institutions of Hainan Province, an empirical research is conducted. A questionnaire survey was carried out in aspects of satisfaction of teaching and scientific research, professional quality, work saturation and teacher’s job satisfaction. The research results shows that the professional quality of teachers of common required courses cannot significantly influence the satisfaction of teaching and scientific research; Job saturation can positively influence teaching and research satisfaction; teaching and research satisfaction can positively influence teachers’ job satisfaction.

With the development of information technology, the demand for information security is increasing. For more convenient and safer needs, the encryption technology based on biometrics has developed rapidly. Among them, iris technology has become an important research object of information security research due to the stability of iris characteristics and its difficulty in forgery. In this paper, the iris feature encryption technology based on the iris is studied by using the method of deep learning as the feature classification method and the iris feature as the research object. The simulation experiment is carried out by using the common iris database. The results show that the method can greatly improve the consistency of iris encryption and improve the security of encryption and decryption process.

The complete chloroplast (cp) genome of Rosa pricei Hayata has been characterized by reference-based assembly using Illumina paired-end data. The complete cp genome is 156,599 base pairs (bp) in le...

The present study was conducted to investigate the effects of dietary acidolysis-oxidized konjac glucomannan (A-OKGM) (0%, 0.4%, 0.8%, and 1.6%) supplementation on the immunity and expression of immune-related genes in Schizothorax prenanti. After feeding for eight weeks, the serum and guts were used for measurement of biochemical parameters, and immune-related gene expression in the gut were also analyzed by real-time quantitative polymerase chain reaction (RT-qPCR). C-reactive protein and IgM levels were significantly higher in the A-OKGM fed groups than in the control group, regardless of the dosage. The 0.4% and 1.6% A-OKGM groups showed significant up-regulation of tumor necrosis factor α (TNFα) in the anterior gut. The 0.8% and 1.6% A-OKGM groups also showed significantly enhanced TNFα expression in the mid- and distal guts. Interleukin-1β (IL-1β) expression in the anterior gut of fish fed with 0.4% and 1.6% A-OKGM diets was significantly enhanced. The 0.8% and 1.6% A-OKGM diets resulted in significantly increased the expression of IL-1β in the distal gut. Similarly, the interleukin-6 (IL-6) messenger RNA (mRNA) levels in the 0.4% and 1.6% diet groups were significantly higher in the anterior gut. The 0.8% and 1.6% A-OKGM diet groups showed significant induction of IL-6 gene expression in the distal gut. A-OKGM modified from KGM can act as an immunostimulant to enhance the immunity of S. prenanti.

Credit scoring process, the most important part in credit risk management, aims at estimating the probability that an applicant will perform bad credit behaviors (e.g., loan default). Managing and developing effective and reliable risk assessment procedures in order to mitigate potential loss caused by new applicants heavily relies on the performance of scoring process. Traditionally this process is manually developed, which is time-consuming. In this paper, we propose an automated credit risk management process based on machine learning to ease the scoring process in order to reduce the human effort. This process is data driven: it leverages machine learning to automatically analyze vast amounts of historical data and build predictive model. We evaluate our process with a real-world proprietary dataset and achieved good performance, which shows the feasibility of using machine learning to facilitate the credit risk management process.