An Adaptive Bounded-Confidence Model of Opinion Dynamics on Networks

Individuals who interact with each other in social networks often exchange ideas and influence each other's opinions. A popular approach to studying the dynamics of opinion spread on networks is by examining bounded-confidence (BC) models, in which the nodes of a network have continuous-valued states that encode their opinions and are receptive to other opinions if they lie within some confidence bound of their own opinion. We extend the Deffuant--Weisbuch (DW) model, which is a well-known BC model, by studying opinion dynamics that coevolve with network structure. We propose an adaptive variant of the DW model in which the nodes of a network can (1) alter their opinion when they interact with a neighboring node and (2) break a connection with a neighbor based on an opinion tolerance threshold and then form a new connection to a node following the principle of homophily. This opinion tolerance threshold acts as a threshold to determine if the opinions of adjacent nodes are sufficiently different to be viewed as discordant. We find that our adaptive BC model requires a larger confidence bound than the standard DW model for the nodes of a network to achieve a consensus. Interestingly, our model includes regions with `pseudo-consensus' steady states, in which there exist two subclusters within an opinion-consensus group that deviate from each other by a small amount. We conduct extensive numerical simulations of our adaptive BC model and examine the importance of early-time dynamics and nodes with initial moderate opinions for achieving consensus. We also examine the effects of coevolution on the convergence time of the dynamics.

Dependency of regret on accuracy of variance estimation for different versions of UCB strategy for Gaussian multi-armed bandits

We consider two variations of upper confidence bound strategy for Gaussian two-armed bandits. Rewards for the arms are assumed to have unknown expected values and unknown variances. It is demonstrated that expected regret values for both discussed strategies are continuous functions of reward variance. A set of Monte-Carlo simulations was performed to show the nature of the relation between variance estimation and losses. It is shown that the regret grows only slightly when the estimation error is fairly large, which allows to estimate the variance during the initial steps of the control and stop this estimation later.

The Impact of Aspiration and Corticosteroid Injection after ACL Injury on Post-Reconstruction Infection Rate (161)

Objectives: Knee injections of bioactive substances, including corticosteroids, hyaluronic acid, and platelet-rich-plasma, are very common in orthopaedic practice. Recently, injecting a corticosteroid into the knee after an anterior cruciate ligament (ACL) injury has been shown in a pilot randomized controlled trial to mitigate articular cartilage damage from the pro-inflammatory effects of the hemearthrosis.1 It is imperative to know if injecting a corticosteroid after ACL tear increases the risk of infection after ACL reconstruction (ACLR). The objective of this study is to report the infection rate in a retrospective cohort of primary bone-tendon-bone (BTB) ACLR patients of one fellowship-trained sports medicine orthopaedic surgeon’s practice where post-injury aspiration and corticosteroid injection occurred prior to ACLR. 1. Lattermann, C. et al. A Multicenter Study of Early Anti-inflammatory Treatment in Patients With Acute Anterior Cruciate Ligament Tear. The American Journal of Sports Medicine 45, 325–333 (2017). Methods: All patients from the ages of 10-65 who underwent primary BTB autograft ACLR by one fellowship-trained sports medicine orthopaedic surgeon between 1/1/2011 and 3/1/2019 at two institutions were reviewed. The variables reviewed were if there was a postoperative infection (as defined by undergoing an intra-articular irrigation and debridement reoperation), if there were any positive cultures, and the medications of the post-injury intra-articular injection (corticosteroid).The time between the following events was also recorded: initial injury and initial presentation, initial injury and corticosteroid injection, injection and ACLR, and ACLR and last date of follow-up. Statistical analysis determined the upper 95% confidence bound for infection probability for the three main groups of the study: the entire cohort, the cohort who underwent post-injury preoperative aspiration and injection, and the cohort who did not undergo aspiration and injection. This statistical approach was taken to determine with 95% confidence the upper limit of what the infection risk would likely be in each group. Results: There were 518 primary BTB ACLR performed with follow-up on 79% (410/518). 174 were found to have undergone a post-injury aspiration and injection, leaving 236 who did not. There were no infections (washout reoperations or positive cultures) in the entire 410 case group. The upper 95% confidence bound for the probability of a postoperative infection is shown in three left columns in table 1 was 0.7% for the whole cohort (n= 410), 1.7% for the cohort who underwent aspiration and injection (n=174), and 1.3% for the cohort that did not (n= 236). Table 1 compares our study to the ACLR infection rate from MOON ADDIN ZOTERO_ITEM CSL_CITATION{"citationID":"XkJTnLOw","properties":{"formattedCitation":"\\super2\\nosupersub{}","plainCitation":"2","noteIndex":0},"citationItems":[{"id":2186,"uris":["http://zotero.org/users/2554704/items/N87NRZ9H"],"uri":["http://zotero.org/users/2554704/items/N87NRZ9H"],"itemData":{"id":2186,"type":"article-journal","container-title":"TheJournal of Bone and Joint Surgery","DOI":"10.2106/JBJS.N.00694","ISSN":"0021-9355","issue":"6","language":"en","page":"450-454","source":"Crossref","title":"FactorsAssociated with Infection Following Anterior Cruciate LigamentReconstruction:","title-short":"Factors Associated withInfection Following Anterior Cruciate Ligament Reconstruction","volume":"97","author":[{"family":"Brophy","given":"RobertH."},{"family":"Wright","given":"RickW."},{"family":"Huston","given":"LauraJ."},{"family":"Nwosu","given":"SamuelK."},{"family":"Spindler","given":"KurtP."},{"family":"Kaeding","given":"ChristopherC."},{"family":"Parker","given":"RichardD."},{"family":"Andrish","given":"JackT."},{"family":"Marx","given":"RobertG."},{"family":"Amendola","given":"Annunziato"},{"family":"Wolf","given":"BrianR."},{"family":"McCarty","given":"EricC."},{"family":"Dunn","given":"WarrenR."}],"issued":{"date-parts":[["2015",3]]}}}],"schema":"https://github.com/citation-style-language/schema/raw/master/csl-citation.json"}2 . The infection rate for the entire MOON cohort is 0.8% (allograft and autograft), the BTB autograft is 0.3%, and the hamstring autograft is1.3%. In our injection group, the mean time between injury and aspiration and injection was 7.6 days. The mean time elapsed between the date of aspiration and injection and the surgical date was 48.9 days. 2. Brophy, R. H. et al. Factors Associated with Infection Following Anterior Cruciate Ligament Reconstruction: The Journal of Bone and Joint Surgery 97, 450–454 (2015). Conclusions: Our data show that aspiration and injection with corticosteroids following ACL injury does not greatly increase post-operative infection risk. This can be observed when comparing the MOON BTB published rate of 0.3% to the upper limit of infection in our entire cohort, which is 0.7% (see Table 1). The difference in upper 95% confidence limits among our three groups is the result of sample size for each cohort. Even in the worst-case scenario for the preoperative injection group, which has the smallest sample size of 174, the upper 95% CI is 1.7%. This is only 1% higher than the confidence bound calculated for the entire MOON cohort (0.7%). When comparing this finding with the MOON autograft hamstring graft infection rate (1.3%) compared to the BTB infection rate (0.3%), the known 1% difference does not alter practice. The results of this study support a very minimal risk of postoperative infection after ACLR with or without a post-injury aspiration and corticosteroid injection, at most potentially 1%.

Multi-Task Gaussian Process Upper Confidence Bound for Hyperparameter Tuning

In many scientific and engineering applications, Bayesian optimization (BO) is a powerful tool for hyperparameter tuning of a machine learning model, materials design and discovery, etc. BO guides the choice of experiments in a sequential way to find a good combination of design points in as few experiments as possible. It can be formulated as a problem of optimizing a “black-box” function. Different from single-task Bayesian optimization, Multi-task Bayesian optimization is a general method to efficiently optimize multiple different but correlated “black-box” functions. The previous works in Multi-task Bayesian optimization algorithm queries a point to be evaluated for all tasks in each round of search, which is not efficient. For the case where different tasks are correlated, it is not necessary to evaluate all tasks for a given query point. Therefore, the objective of this work is to develop an algorithm for multi-task Bayesian optimization with automatic task selection so that only one task evaluation is needed per query round. Specifically, a new algorithm, namely, multi-task Gaussian process upper confidence bound (MT-GPUCB), is proposed to achieve this objective. The MT-GPUCB is a two-step algorithm, where the first step chooses which query point to evaluate, and the second step automatically selects the most informative task to evaluate. Under the bandit setting, a theoretical analysis is provided to show that our proposed MT-GPUCB is no-regret under some mild conditions. Our proposed algorithm is verified experimentally on a range of synthetic functions as well as real-world problems. The results clearly show the advantages of our query strategy for both design point and task.

Multi-Task Gaussian Process Upper Confidence Bound for Hyperparameter Tuning

In many scientific and engineering applications, Bayesian optimization (BO) is a powerful tool for hyperparameter tuning of a machine learning model, materials design and discovery, etc. BO guides the choice of experiments in a sequential way to find a good combination of design points in as few experiments as possible. It can be formulated as a problem of optimizing a “black-box” function. Different from single-task Bayesian optimization, Multi-task Bayesian optimization is a general method to efficiently optimize multiple different but correlated “black-box” functions. The previous works in Multi-task Bayesian optimization algorithm queries a point to be evaluated for all tasks in each round of search, which is not efficient. For the case where different tasks are correlated, it is not necessary to evaluate all tasks for a given query point. Therefore, the objective of this work is to develop an algorithm for multi-task Bayesian optimization with automatic task selection so that only one task evaluation is needed per query round. Specifically, a new algorithm, namely, multi-task Gaussian process upper confidence bound (MT-GPUCB), is proposed to achieve this objective. The MT-GPUCB is a two-step algorithm, where the first step chooses which query point to evaluate, and the second step automatically selects the most informative task to evaluate. Under the bandit setting, a theoretical analysis is provided to show that our proposed MT-GPUCB is no-regret under some mild conditions. Our proposed algorithm is verified experimentally on a range of synthetic functions as well as real-world problems. The results clearly show the advantages of our query strategy for both design point and task.

Analysis of the specificity of a COVID-19 antigen test in the Slovak mass testing program

Aims Mass antigen testing programs have been challenged because of an alleged insufficient specificity, leading to a large number of false positives. The objective of this study is to derive a lower bound of the specificity of the SD Biosensor Standard Q Ag-Test in large scale practical use. Methods Based on county data from the nationwide tests for SARS-CoV-2 in Slovakia between 31.10.–1.11. 2020 we calculate a lower confidence bound for the specificity. As positive test results were not systematically verified by PCR tests, we base the lower bound on a worst case assumption, assuming all positives to be false positives. Results 3,625,332 persons from 79 counties were tested. The lowest positivity rate was observed in the county of Rožňava where 100 out of 34307 (0.29%) tests were positive. This implies a test specificity of at least 99.6% (97.5% one-sided lower confidence bound, adjusted for multiplicity). Conclusion The obtained lower bound suggests a higher specificity compared to earlier studies in spite of the underlying worst case assumption and the application in a mass testing setting. The actual specificity is expected to exceed 99.6% if the prevalence in the respective regions was non-negligible at the time of testing. To our knowledge, this estimate constitutes the first bound obtained from large scale practical use of an antigen test.

Uncertainty Assessment of Entropy-Based Circular Channel Shear Stress Prediction Models Using a Novel Method

Entropy models have been recently adopted in many studies to evaluate the shear stress distribution in open-channel flows. Although the uncertainty of Shannon and Tsallis entropy models were analyzed separately in previous studies, the uncertainty of other entropy models and comparisons of their reliability remain an open question. In this study, a new method is presented to evaluate the uncertainty of four entropy models, Shannon, Shannon-Power Law (PL), Tsallis, and Renyi, in shear stress prediction of the circular channels. In the previous method, the model with the largest value of the percentage of observed data within the confidence bound (Nin) and the smallest value of Forecasting Range of Error Estimation (FREE) is the most reliable. Based on the new method, using the effect of Optimized Forecasting Range of Error Estimation (FREEopt) and Optimized Confidence Bound (OCB), a new statistic index called FREEopt-based OCB (FOCB) is introduced. The lower the value of FOCB, the more certain the model. Shannon and Shannon PL entropies had close values of the FOCB equal to 8.781 and 9.808, respectively, and had the highest certainty, followed by ρgRs and Tsallis models with close values of 14.491 and 14.895, respectively. However, Renyi entropy, with the value of FOCB equal to 57.726, had less certainty.

Collaborative Filtering-Based Recommendation System Using Time Decay Model

Online collaborative movie recommendation systems attempt to help customers accessing their favourable movies by gathering exactly comparable neighbors between the movies from their chronological identical ratings. Collaborative filtering-based movie recommendation systems require viewer-specific data, and the need for collecting viewer-specific data diminishes the effectiveness of the recommendation. To solve this problem, the authors employ an effective multi-armed bandit called upper confidence bound, which is applied to automatically recommend the movies for the users. In addition, the concept of time decay is provided in a mathematical definition that redefines the dynamic item-to-item similarity. Then, two patterns of time decay are analyzed, namely concave and convex functions, for simulation. The experiment test the MovieLens 100K dataset. The proposed method attains a maximum F-measure of 98.45 whereas the existing method reaches a minimum F-measure of only 95.60. The presented model adaptively responds to new users, can provide a better service, and generate more user engagement.