Search

Mean reversion is a financial term used to assume that the price of an asset will tend to converge to the average price over time. Implementation of mean reversion as a timing strategy involves both identifying the trading range of a security and calculating average prices using quantitative methods. Mean-reverting strategies apply the classic idea of selling high and buying low in the stock market, which generally performs well in volatile markets. Based on empirical analysis, this paper calculates the value of Bilibili, iQiyi and YouTube, compares the deviation index and return rate of the three companies. Subsequently, we use the deviation multiple and average return rate of historical statistics to judge the entry time. The significance of this study is that it provides a reference for Internet video companies to choose the path of mean reversion, and also provides an investment strategy for investors. These results shed light on guiding further applications of mean-revision strategy for different underlying assets.

To better understand the benzene alkylation with chloroaluminate ionic liquids (ILs) as catalyst, the interfacial properties between the benzene/butene binary reactants and chloroaluminate ILs with varying cation alky chain length and different anions were investigated using molecular dynamics (MD) simulations. The results indicate that ILs can obviously improve the interfacial width, solubility and diffusion of reactants compared to H2SO4. The longer alkyl chains of cations present a density enrichment at the interface and protrude into the binary reactants phase. Furthermore, the ILs consisting of 1-octyl-3-methylimidazolium cations ([Omim]+) and the stronger acidity heptachlorodialuminateanions ([Al2Cl7]–) are more beneficial to promote the interfacial width and facilitate the dissolution and diffusion of benzene in both the IL bulk and the interfacial region in comparison to the ones with shorter alkyl chains cations and weaker acidity anions. The information gives us a better guideline for the design of ILs for benzene alkylation.

This work is under review of IEEE Transactions on Parallel and Distributed Systems.

Purpose: To observe retinal sensitivity (RS) and microvasculature changes in preoperative and postoperative idiopathic macular epiretinal membrane (iERM) with optical coherence tomography (OCT) classification by microperimetry (MP) and optical coherence tomography angiography (OCTA). Methods: In this retrospective study, 53 eyes of 53 patients affected with iERM were enrolled and 35 iERM eyes underwent pars plana vitrectomy associated with internal limiting membrane peeling. All iERM eyes were evaluated and graded by OCT. According to the thickness of the fovea relative to the surrounding macula from OCT radial line scans, we classified iERM into 3 grades. MP was used to measure RS of macular within 20° centered on the fovea and fixation stability (FS). OCTA was used to measure the superficial and deep capillary plexus (DCP) layers using 6 mm × 6 mm scans. Measurements were taken at the baseline visit and 3-month postoperation. Evaluated factors included: best corrected visual acuity, RS, FS, vessel density. Results: Eyes with iERM had worse RS than control eyes (p < 0.0001), especially in higher grade iERM eyes. And higher grade iERM eyes had a marked increase foveal vessel density (FVD) (p < 0.0001) in superficial capillary plexus (SCP) and DCP, and a significant decrease parafoveal vessel density (Para VD) in DCP. After surgery, there was no difference in RS and FS than preoperation (p > 0.05), whether in grade 1 or 2 iERM eyes. However, in OCTA analyses, grade 1 iERM eyes had less whole vessel density (WVD) and perifoveal vessel density (Peri VD) in SCP, than preoperation (p < 0.05). And grade 1 iERM eyes had higher Para VD and Peri VD in DCP (p < 0.05), than preoperative OCTA results. Grade 2 iERM eyes had less WVD, FVD, and Peri VD in SCP, than preoperation (p < 0.05). After surgery, in grade 2 iERM eyes, there was a positive correlation between whole retinal sensitivity and WVD (p < 0.05) in SCP. And there were positive correlations between Para RS and Para VD (p < 0.05), Peri RS and Peri VD (p < 0.05) in SCP. Conclusions: Classification helps to reveal different changes in iERM eyes. Higher grade iERM eyes had worse RS and microvasculature changes. MP and OCTA play an important role in assessing iERM and help surgical decisions for patients with iERM.

Purpose: To quantitatively assess the retinal capillary vessel density (VD) and foveal avascular zone (FAZ) area in acute and chronic central serous chorioretinopathy (CSC) patients compared to healthy subjects using optical coherence tomography angiography (OCTA) and determine their correlation with visual acuity (VA). Methods: This was a cross-sectional, case-control study. All subjects were divided into three groups: acute CSC, chronic CSC, and healthy controls. The best-corrected VA test, optical coherence tomography, and OCTA were performed on the same day. In this study, we recorded and compared the main data of FAZ and VD in 3 × 3 mm OCTA images among three groups, also calculated the correlation between OCTA variables and VA in CSC patients. Results: A total of 148 subjects (148 eyes) were included in this study, with 50 eyes in acute CSC, 48 eyes in chronic CSC, and 50 eyes in control groups. The mean age was 50.11 ± 8.14 years. Chronic CSC had the significantly lowest VD both on the superficial and deep retina and the largest FAZ (0.39 ± 0.13 mm2) compared to acute CSC and controls (all p < 0.05). There was no statistical difference between acute CSC and controls on VD or FAZ (both p > 0.05). In chronic CSC, both the decreased VD on deep retina and expanded FAZ were correlated with a worse VA (both p < 0.05). Conclusions: OCTA could serve as a reliable tool to evaluate different courses of CSC quantitively. In 3 × 3 mm OCTA images, both decreased VD and expanded FAZ were observed in chronic CSC patients.

The aim of this study was to compare intraocular cytokines among different types of pigment epithelial detachments (PEDs) in patients with chronic central serous chorioretinopathy (CSC) and to investigate the association of cytokine levels and PED types with response to anti-vascular endothelial growth factor (VEGF) therapy.We included 88 patients with chronic CSC and 30 controls. The anti-VEGF agent conbercept was given intravitreally to chronic CSC patients. Cytokines VEGF, interleukin-6 (IL-6), IL-8, IL-10, interferon-inducible protein-10 and monocyte chemoattractant protein-1 in aqueous humour were measured. Treatment efficacy, cytokine levels, changes in best-corrected visual acuity (BCVA) and optical coherence tomography parameters were assessed at baseline and 1 month after treatment.Patients were divided into three groups: flat irregular PED (FIPED) with choroidal neovascularization (CNV), FIPED without CNV and focal PED. Vascular endothelial growth factor (VEGF) was the only cytokine significantly higher in chronic CSC FIPED patients. There were no significant differences in VEGF between FIPED patients with or without CNV (p = 0.234). At 1 month after conbercept injection, treatment effective rates in FIPED patients with or without CNV were significantly higher than in patients with focal PED (p 0.05). Best-corrected visual acuity (BCVA) was improved in both FIPED groups (p 0.05), but not in the focal PED group (p = 0.180). All three groups had significant decreases in central macular thickness (p 0.05), and PED heights in FIPED patients were reduced (p 0.05).Patients with FIPED in chronic CSC had elevated intraocular VEGF levels and responded favourably to conbercept. Anti-VEGF treatment may be an option for FIPED CSC patients with or without secondary CNV.

AIM: To analyse macular microvascular alterations in myopic choroidal neovascularization (mCNV) and the efficiency of anti-vascular endothelial growth factor (anti-VEGF) therapy for mCNV by optical coherence tomography angiography (OCTA). METHODS: A total of 123 patients were included in this retrospective study, divided into mCNV group, high myopia (HM) group, and normal group at the Affiliated Eye Hospital of Wenzhou Medical University from January 2017 to January 2019. Superficial vessel density, deep capillary density, foveal avascular zone (FAZ) area, A-circularity index (AI) and vessel density around the 300 μm width of the FAZ region density (FD) and the area of choroidal neovascularization (CNV) lesion (only for mCNV group) were measured on 3×3 mm2 OCTA images. FAZ area was corrected for axial length. Central macular thickness (CMT) was measured on OCT in mCNV group. Compared the parameters on OCTA of 3 groups and pre-anti-VEGF and post-anti-VEGF at 1, 2, 3, and 6mo follow-up in mCNV group. RESULTS: There were significant differences among 3 groups in superficial vessel density, deep capillary density and FD (P

Lots of studies focus on distributed traffic engineering (TE) where routers make routing decisions independently. Existing approaches usually tackle distributed TE problems through traditional optimization methods. However, due to the intrinsic complexity of the distributed TE problems, routing decisions cannot be obtained efficiently, which leads to significant performance degradation, especially for highly dynamic traffic. Emerging machine learning technologies like deep reinforcement learning (DRL) provide a new choice to address TE problems in an experience-driven method. In this paper, we propose DATE, a distributed and adaptive TE framework with DRL. DATE distributes well-trained agents to the routers in the located network. Each agent makes local routing decisions independently based on link utilization ratios flooded by each router periodically. To coordinate the distributed agents to achieve the global optimization in different traffic conditions, we construct candidate paths, develop the agents carefully, and realize a virtual environment to train the agents with a DRL algorithm. We do extensive simulations and experiments using real-world network topologies with both real and synthetic traffic traces. The results show that DATE outperforms some existing approaches and yields near-optimal performance with superior robustness.

AIM: To compare the safety and efficacy of conbercept intravitreal injection and half-dose photodynamic therapy (PDT) in treating chronic central serous chorioretinopathy (CSC). METHODS: This study was retrospective. Thirty-seven patients (37 eyes) with chronic CSC received conbercept injections while 57 patients (57 eyes) were treated with half-dose PDT. All subjects were followed in 6mo. Outcome measures included change in best-corrected visual acuity (BCVA), central macular thickness (CMT), subfoveal choroidal thickness (SFCT), and resolution of subretinal fluid (SRF). RESULTS: There was no adverse event observed in either treatment group. At the 6-month follow-up, 26 eyes (70.3%) in the conbercept group and 54 eyes (94.7%) in the half-dose PDT group (P0.05). The SFCT also improved in all subjects (P

Emerging applications raise critical QoS requirements for the Internet. The improvements of flow classification technologies, software defined networks (SDN), and programmable network devices make it possible to fast identify users’ requirements and control the routing for fine-grained traffic flows. Meanwhile, the problem of optimizing the forwarding paths for traffic flows with multiple QoS requirements in an online fashion is not addressed sufficiently. To address the problem, we propose DRL-OR, an online routing algorithm using multi-agent deep reinforcement learning. DRL-OR organizes the agents to generate routes in a hop-by-hop manner, which inherently has good scalability. It adopts a comprehensive reward function, an efficient learning algorithm, and a novel deep neural network structure to learn an appropriate routing policy for different types of flow requirements. To guarantee the reliability and accelerate the online learning process, we further introduce safe learning mechanism to DRL-OR. We implement DRL-OR under SDN architecture and conduct Mininet-based experiments by using real network topologies and traffic traces. The results validate that DRL-OR can well satisfy the requirements of latency-sensitive, throughput-sensitive, latency-throughput-sensitive, and latency-loss-sensitive flows at the same time, while exhibiting great adaptiveness and reliability under the scenarios of link failure, traffic change, and partial deployment.

Transportation is an important link in the mining process, and large trucks are one of the important tools for mine transportation. Due to their large size and small driving position, large trucks have a blind spot, which is a hidden danger to the safe transportation of mines and has a great impact on production efficiency and economic loss. The traditional large truck safety warning system mainly uses the ultrasonic short-distance ranging method, radar ranging method, GPS (Global Positioning System) technology, and so on. The disadvantage of these methods is that they are affected by the environment and weather, and they cannot display the object status in real time. Therefore, it is becoming increasingly important to realize the large truck safety warning system based on machine vision. Therefore, this paper proposes a lightweight SSD (Single Shot MultiBox Detector) model and an atrous convolution to build a large-truck object recognition model. First, the training images are collected and marked. Then, the object recognition model is established by using the lightweight SSD model. The atrous convolutional layer is introduced to improve small object detection accuracy. In the end, the objectness prior method is used to improve the classification speed. Experimental results show that, compared with the original SSD model, the lightweight SSD model occupies less space and runs faster. The lightweight SSD model with the atrous convolutional layer is more sensitive to small objects and improves detection accuracy. The objectness prior method further improves the identification speed. Compared with the traditional large truck safety warning, the system is not affected by the environment and realizes the visualization of large truck safety warning.

Purpose. To evaluate the efficacy and safety of conbercept for patients with chronic central serous chorioretinopathy (CSC). Methods. A retrospective clinical study. Thirty-one patients (35 eyes) with chronic CSC were given intravitreal injections of conbercept and followed up for at least 6 months. Observed indicators included best-corrected visual acuity (BCVA), central macular thickness (CMT), and resolution of subretinal fluid (SRF). Serial changes in BCVA and CMT were analyzed by using repeated measures analysis of variance. Results. During the 6-month follow-up, the mean number of injections required and performed was 1.77 ± 0.60. The logMAR BCVA was 0.48 ± 0.26 at the baseline, 0.34 ± 0.26, 0.30 ± 0.26, 0.27 ± 0.26, 0.24 ± 0.26, and 0.23 ± 0.26 at 2-week and 1-, 2-, 3-, and 6-month follow-ups, respectively (F = 27.173, P<0.05). CMT was 313.74 ± 144.51 μm at the baseline and decreased to 263.49 ± 120.44 μm, 225.91 ± 91.98 μm, 195.77 ± 66.69 μm, 189.74 ± 65.41 μm, and 199.49 ± 81.50 μm at 2-week and 1-, 2-, 3-, and 6-month follow-ups, respectively (F = 18.093, P<0.05). Full resolution of SRF was achieved in 8 (22.9%) eyes at 1 month, 16 (45.7%) eyes at 2 months, 22 (62.9%) eyes at 3 months, and 27 (77.1%) eyes at 6 months after the initial treatment of anti-VEGF injection. No severe adverse event was noted relevant to the therapy. Conclusions. Intravitreal injection of conbercept may effectively reduce the CMT and improve the BCVA in chronic CSC in a short term of 6 months.

Hydrometallurgy is a metallurgical method for processing complex ores and low-grade ores while reducing environmental pollution. The density of the thickening process in hydrometallurgical production is rather poor, and there are many interference factors, resulting in frequent failures in the density of the thickening process. The main focus of this paper is to propose a method of fault monitoring and diagnosis for the density of the thickening process in hydrometallurgy. First, through the support vector machine (SVM) algorithm, the fault detection model is established to monitor the blockage of the underflow pipeline of the thickener. Second, the fault diagnosis model is established by using the random forest algorithm, and particle swarm optimization is used to optimize the fault diagnosis model. The fault type is judged using the optimized diagnosis model, and the corresponding treatment measures are taken accordingly.

PURPOSE In eyes with diabetic macular oedema (DME), aqueous humour (AH) cytokine levels before and after anti-vascular endothelial growth factor (VEGF) treatment were compared and correlated with optical coherence tomography structural parameters. METHODS This prospective study included 56 control patients with cataracts and 83 patients with DME manifesting as diffuse retinal thickening (DRT), cystoid macular oedema and serous retinal detachment (SRD). AH samples were obtained before intravitreal injection of anti-VEGF or cataract surgery. VEGF, interleukin (IL)-6, IL-8, IL-10, interferon-inducible protein 10 (IP-10) and monocyte chemotactic protein 1 (MCP-1) levels were measured by multiplex bead assay. AH cytokine levels, central macular thickness (CMT), number of hyper-reflective foci (HF), continuity of external limiting membrane and ellipsoid zone (EZ) and best-corrected visual acuity were evaluated. RESULTS In SRD, IL-6 and MCP-1 levels and HF were increased (all p

In many real-world tasks, a team of learning agents must ensure that their optimized policies collectively satisfy required peak and average constraints, while acting in a decentralized manner. In this paper, we consider the problem of multi-agent reinforcement learning for a constrained, partially observable Markov decision process – where the agents need to maximize a global reward function subject to both peak and average constraints. We propose a novel algorithm, CMIX, to enable centralized training and decentralized execution (CTDE) under those constraints. In particular, CMIX amends the reward function to take peak constraint violations into account and then transforms the resulting problem under average constraints to a max-min optimization problem. We leverage the value function factorization method to develop a CTDE algorithm for solving the max-min optimization problem, and two gap loss functions are proposed to eliminate the bias of learned solutions. We evaluate our CMIX algorithm on a blocker game with travel cost and a large-scale vehicular network routing problem. The results show that CMIX outperforms existing algorithms including IQL, VDN, and QMIX, in that it optimizes the global reward objective while satisfying both peak and average constraints. To the best of our knowledge, this is the first proposal of a CTDE learning algorithm subject to both peak and average constraints.

Glycerol is an abundant byproduct of biodiesel production that has significant industrial value and can be converted into dihydroxyacetone (DHA). DHA is widely used for the production of various chemicals, pharmaceuticals, and food additives. Gluconobacter can convert glycerol to DHA through two different pathways, including membrane-bound dehydrogenases with pyrroloquinoline quinone (PQQ) and NAD(P)+ -dependent enzymes. Previous work has indicated that membrane-bound dehydrogenases are present in Gluconobacter oxydans and Gluconobacter frateurii, but the metabolic mechanism of Gluconobacter thailandicus's glycerol conversion is still not clear. Through in-depth analysis of the G. thailandicus genome and annotation of its metabolic pathways, we revealed the existence of both PQQ and NAD(P)+ -dependent enzymes in G. thailandicus. In addition, this study provides important information related to the tricarboxylic acid cycle, glycerol dehydrogenase level, and phylogenetic relationships of this important species.

Traffic engineering (TE) attracts extensive researches over the years. Operators expect to design a TE scheme which accommodates traffic dynamics well and achieves good TE performance with little overhead. Some approaches like oblivious routing compute an optimal static routing based on a large traffic matrix (TM) range, which usually leads to much performance loss. Many approaches compute routings based on one or a few representative TMs obtained from observed historical TMs. However, they may suffer performance degradation for unexpected TMs and usually induce much overhead of system operating. In this paper, we propose ALTE, an adaptive and low-cost TE scheme based on TM classification. We develop a novel clustering algorithm to properly group a set of historical TMs into several clusters and compute a candidate routing for each TM cluster. A machine learning classifier is trained to infer the proper candidate routing online based on the features extracted from some easily measured statistics. We implement a system prototype of ALTE and do extensive simulations and experiments using both real and synthetic traffic traces. The results show that ALTE achieves near-optimal performance for dynamic traffic and introduces small overhead of routing updates.

Purpose. The objective of this study was to evaluate central corneal thickness (CCT) in Chinese premature infants at different postnatal stages to study the peak point and analyze influential factors on CCT development. Methods. This was a cross-sectional study of premature infants. Initial CCT measurement was taken at 34 weeks of gestational age (GA) and at intervals until 88 weeks of postmenstrual age (PMA) was reached. The comparison and correlation analysis were carried out to access the association of CCT with gender, birth weight (BW), GA, and retinopathy of prematurity (ROP) for each PMA. The premature infants were divided into the thick CCT group and the thin CCT group according to the average CCT at 40 w. And the difference in CCT between the two groups at subsequent 52 w and 64 w was compared. Results. A total of 1726 premature infants (3463 measurements) with an average of 2.21 ± 1.57 measurements were included in this study. The CCT decreased from 34 w GA to 52 w PMA (R = 92.36, P<0.0001) and then reached a plateau (R = 2.541, P=0.3567). Male P<0.05, low BW P<0.05, and low GA P<0.05 were associated with thicker CCT at the early stage of PMA. The premature infants who had thick CCT at 40 w would have thick CCT at 52 w and 64 w accordingly. Conclusions. The CCT values of premature infants decreased over time and plateaued at 52 w PMA. Gender, BW, and GA were considered as the influential factors of CCT at the early stage of PMA. Moreover, CCT at 40 w could forecast its development trend at 52 w or 64 w after birth.