Bayesian Program Research Articles

Multivariate Bayesian Analyses in Nursing Research: An Introductory Guide.

In the era of precision health, nursing research has increasingly focused on the analysis of large, multidimensional data sets containing multiple correlated phenotypes (e.g., symptoms). This presents challenges for statistical analyses, especially in genetic association studies. For example, the inclusion of multiple symptoms within a single model can raise concerns about multicollinearity, while individual SNP-symptom analyses may obscure complex relationships. As such, many traditional statistical approaches often fall short in providing a comprehensive understanding of the complexity inherent in many nursing-focused research questions. Multivariate Bayesian approaches offer the unique advantage of allowing researchers to ask questions that are not feasible with traditional approaches. Specifically, these methods support the simultaneous exploration of multiple phenotypes, accounting for the underlying correlational structure between variables, and allow for formal incorporation of existing knowledge into the statistical model. By doing so, they may provide a more realistic view of statistical relationships within a biological system, potentially uncovering new insights into well-established and undiscovered connections, such as the probabilities of association and direct versus indirect effects. This valuable information can help us better understand our phenotypes of interest, leading to more effective nurse-led intervention and prevention programs. To illustrate these concepts, this paper includes an application section covering two specific multivariate Bayesian analysis software programs, bnlearn and mvBIMBAM, with an emphasis on interpretation and extension to nursing research. To complement the paper, we provide access to a detailed online tutorial, including executable R code and a synthetic data set, so the concepts can be more easily extended to other research questions.

Automated calculation and reporting of vancomycin area under the concentration-time curve: a simplified single-trough concentration-based equation approach.

Vancomycin, a crucial antibiotic for Gram-positive bacterial infections, requires therapeutic drug monitoring (TDM). Contemporary guidelines advocate for AUC-based monitoring; however, using Bayesian programs for AUC estimation poses challenges. We aimed to develop and evaluate a simplified AUC estimation equation using a steady-state trough concentration (Ctrough) value. Utilizing 1,034 TDM records from 580 general hospitalized patients at a university-affiliated hospital in Ulsan, we created an equation named SSTA that calculates the AUC by applying Ctrough, body weight, and single dose as input variables. External validation included 326 records from 163 patients at a university-affiliated hospital in Seoul (EWUSH) and literature data from 20 patients at a university-affiliated hospital in Bangkok (MUSI). It was compared with other AUC estimation models based on the Ctrough, including a linear regression model (LR), a sophisticated model based on the first-order equation (VancoPK), and a Bayesian model (BSCt). Evaluation metrics, such as median absolute percentage error (MdAPE) and the percentage of observations within ±20% error (P20), were calculated. External validation using the EWUSH data set showed that SSTA, LR, VancoPK, and BSCt had MdAPE values of 6.4, 10.1, 6.6, and 7.5% and P20 values of 87.1, 82.5, 87.7, and 83.4%, respectively. External validation using the MUSI data set showed that SSTA, LR, and VancoPK had MdAPEs of 5.2, 9.4, and 7.2%, and P20 of 95, 90, and 95%, respectively. Owing to its decent AUC prediction performance, simplicity, and convenience for automated calculation and reporting, SSTA could be used as an adjunctive tool for the AUC-based TDM.

PhyloBench: A Benchmark for Evaluating Phylogenetic Programs.

Phylogenetic inference based on protein sequence alignment is a widely used procedure. Numerous phylogenetic algorithms have been developed, most of which have many parameters and options. Choosing a program, options, and parameters can be a nontrivial task. No benchmark for comparison of phylogenetic programs on real protein sequences was publicly available. We have developed PhyloBench, a benchmark for evaluating the quality of phylogenetic inference, and used it to test a number of popular phylogenetic programs. PhyloBench is based on natural, not simulated, protein sequences of orthologous evolutionary domains. The measure of accuracy of an inferred tree is its distance to the corresponding species tree. A number of tree-to-tree distance measures were tested. The most reliable results were obtained using the Robinson-Foulds distance. Our results confirmed recent findings that distance methods are more accurate than maximum likelihood (ML) and maximum parsimony. We tested the bayesian program MrBayes on natural protein sequences and found that, on our datasets, it performs better than ML, but worse than distance methods. Of the methods we tested, the Balanced Minimum Evolution method implemented in FastME yielded the best results on our material. Alignments and reference species trees are available at https://mouse.belozersky.msu.ru/tools/phylobench/ together with a web-interface that allows for a semi-automatic comparison of a user's method with a number of popular programs.

Development and validation of a Bayesian unfolding program using OpenBUGS open-source software for Bonner spheres neutron spectrometry

This study presents the development of a novel unfolding program based on a Bayesian method for Bonner spheres spectrometry. The program incorporates a priori information on the neutron spectrum, employing a linear superposition of three non-negative parametric functions: thermal-Maxwellian distribution, an intermediate region represented by a straight line in lethargy space (1/E behavior), and the Watt fission distribution for the fast region. The effectiveness of the developed program was validated through its application to two reference cases from the EURADOS 2018 exercise (LINAC and Skyshine), followed by its implementation in characterizing the neutron field around an OB26 neutron irradiator at the Secondary Standards Dosimetry Laboratory (SSDL) of the Nuclear Research Center of Algiers (CRNA), Algeria. While validation results for the LINAC case were promising, achieving comparable neutron spectrum shape and integral quantities to the reference, challenges were encountered with the Skyshine case. However, these challenges were partially addressed through sensitivity analysis, resulting in slightly improved outcomes compared to the previous participation and to the reference. Application of the program to characterize the neutron field around the OB26 irradiator yielded satisfactory results when compared to reference data. A two-step sensitivity analysis further enhanced the outcomes, initially substituting the Watt fission distribution with the evaporation distribution, thereby improving spectrum shape and integral quantities. Subsequently, utilizing the obtained neutron spectra as prior information for unfolding codes MAXED and GRAVEL ensured stability and agreement with reference data and preliminary results.

Learning Abstraction Selection for Bayesian Program Analysis

We propose a learning-based approach to select abstractions for Bayesian program analysis. Bayesian program analysis converts a program analysis into a Bayesian model by attaching probabilities to analysis rules. It computes probabilities of analysis results and can update them by learning from user feedback, test runs, and other information. Its abstraction heavily affects how well it learns from such information. There exists a long line of works in selecting abstractions for conventional program analysis but they are not effective for Bayesian program analysis. This is because they do not optimize for generalization ability. We propose a data-driven framework to solve this problem by learning from labeled programs. Starting from an abstraction, it decides how to change the abstraction based on analysis derivations. To be general, it considers graph properties of analysis derivations; to be effective, it considers the derivations before and after changing the abstraction. We demonstrate the effectiveness of our approach using a datarace analysis and a thread-escape analysis.

Comparison of vancomycin area under the curve calculated based on Bayesian approach versus equation-based approach.

Area under the curve (AUC)-based vancomycin dose adjustment is recommended to treat methicillin-resistant Staphylococcus aureus (MRSA) infections. AUC estimation methods include Bayesian software programs and simple analytical equations. This study compared the AUC obtained using the Bayesian approach with that obtained using an equation-based approach. Patients receiving intravenous vancomycin for MRSA infection were included. Peak and trough levels were measured for each patient on days 3, 7, and 10 post vancomycin dosing (day 1). AUC was calculated using software based on the Bayesian method (MwPharm Online) and an equation-based calculator, Stanford Health Care (SHC) calculator. The AUC estimated using MwPharm Online was similar to that estimated using the SHC calculator. The geometric mean ratio (GMR) and their 90% confidence intervals (90% CI) were 1.08 (1.05-1.11), 1.03 (0.99-1.07), and 0.99 (0.94-1.05) at days 3, 7, and 10, respectively. Furthermore, according to the software used, there were no significant differences in the proportions of patients in the categories "within" and "below or above" the AUC target range. Additionally, trough levels predicted by both software programs were lower than the observed ones. Still, there was no significant difference between the predicted and observed peak levels for both software programs on day 10. AUC calculated using the Bayesian software allows for calculation with samples at a non-steady state, can integrate covariates, and is interconvertible with that estimated using an equation-based calculator, which is simpler and relies on fewer assumptions. Therefore, either method can be used, considering each method's strengths and limitations.

Vancomycin AUC-Based Dosing Practices in a Non-Teaching Community Hospital and Associated Outcomes: A One-Year Survey of Uniform Targets for Infections with or without MRSA.

Intravenous (IV) vancomycin area under the curve (AUC)-based dosing is used uniformly for Gram-positive organisms in non-teaching community hospitals. However, evidence for using vancomycin AUC-based dosing for non-methicillin-resistant Staphylococcus aureus (non-MRSA) and less serious infections is limited in the literature. A gap in the literature also exists with respect to comparisons between the outcomes that can be derived using the regimens suggested by Bayesian programs and target doses of the AUC of 400-499 and 500-600. A retrospective review of all patients hospitalized in a non-teaching community hospital who used AUC-based vancomycin was performed over a 1-year period. Only 17.6% of the included patients had confirmed MRSA. The values for the overall early response rate, 30-day all-cause mortality, and rate of acute kidney injury (AKI) were 50.3%, 11.3%, and 3.8%, respectively, in this population. In regression analysis, compared to non-MRSA infections, a significantly higher rate of early response was seen in patients with MRSA (unadjusted OR = 2.68, 95% CI [1.06-6.76] p = 0.04). Patients in the AUC 400-499 group had a non-significant higher incidence of 30 d mortality and new AKI compared to patients in the AUC 500-600 group. In our Kaplan-Meier survival analysis, there was no statistically significant difference between the comparison groups. Early response was lower in patients with non-MRSA compared to patients with MRSA despite achieving the AUC target. There was no apparent difference in clinical outcomes between the higher and lower AUC groups. Further large-scale research is needed to confirm these findings.

Compositional diversity in visual concept learning

Humans leverage compositionality to efficiently learn new concepts, understanding how familiar parts can combine together to form novel objects. In contrast, popular computer vision models struggle to make the same types of inferences, requiring more data and generalizing less flexibly than people do. Here, we study these distinctively human abilities across a range of different types of visual composition, examining how people classify and generate “alien figures” with rich relational structure. We also develop a Bayesian program induction model which searches for the best programs for generating the candidate visual figures, utilizing a large program space containing different compositional mechanisms and abstractions. In few shot classification tasks, we find that people and the program induction model can make a range of meaningful compositional generalizations, with the model providing a strong account of the experimental data as well as interpretable parameters that reveal human assumptions about the factors invariant to category membership (here, to rotation and changing part attachment). In few shot generation tasks, both people and the models are able to construct compelling novel examples, with people behaving in additional structured ways beyond the model capabilities, e.g. making choices that complete a set or reconfigure existing parts in new ways. To capture these additional behavioral patterns, we develop an alternative model based on neuro-symbolic program induction: this model also composes new concepts from existing parts yet, distinctively, it utilizes neural network modules to capture residual statistical structure. Together, our behavioral and computational findings show how people and models can produce a variety of compositional behavior when classifying and generating visual objects.

Efficient Bayesian inference under the multispecies coalescent with migration.

Analyses of genome sequence data have revealed pervasive interspecific gene flow and enriched our understanding of the role of gene flow in speciation and adaptation. Inference of gene flow using genomic data requires powerful statistical methods. Yet current likelihood-based methods involve heavy computation and are feasible for small datasets only. Here, we implement the multispecies-coalescent-with-migration model in the Bayesian program bpp, which can be used to test for gene flow and estimate migration rates, as well as species divergence times and population sizes. We develop Markov chain Monte Carlo algorithms for efficient sampling from the posterior, enabling the analysis of genome-scale datasets with thousands of loci. Implementation of both introgression and migration models in the same program allows us to test whether gene flow occurred continuously over time or in pulses. Analyses of genomic data from Anopheles mosquitoes demonstrate rich information in typical genomic datasets about the mode and rate of gene flow.

Local Adaptation in Natural Populations of Toona ciliata var. pubescens Is Driven by Precipitation and Temperature: Evidence from Microsatellite Markers

Environmental factors are strong drivers of local adaptation in forest tree species. Toona ciliata var. pubescens, an endangered tree species endemic to China, is widely distributed across Eastern and Southwestern China. In this study, we used 8 genomic microsatellite markers and 17 EST-SSR markers across nine populations from the Yunnan–Kweichow Plateau and Eastern China, to explore the adaptive variation and genetic structure of T. ciliata var. pubescens. Patterns of population structure were apparent using a Bayesian clustering program, STRUCTURE, which identified four distinct clusters. We identified four outlier loci that were potentially under selection using the Dirichlet-multinomial and hierarchic simulation models. Through the Mantel test, it was found that geographic and climatic factors have jointly affected the genetic structure of T. ciliata var. pubescens in the study area. Based on redundancy analysis (RDA), it was shown that the correlation between climatic variables associated with variation is stronger than that of geographic variables. It is worth mentioning that the eight alleles from outlier loci have potentially adaptive and are associated with either precipitation or temperature variables. All analyses revealed high genetic diversity and significant genetic differentiation in the populations of T. ciliata var. pubescens. This indicated that the climatic variables including precipitation and temperature are drivers of local adaptation in the populations of T. ciliata var. pubescens.

문항반응이론에서 문항모수 추정 알고리즘을 위한 C언어 코드 구현에 대한 연구

The purpose of this paper is to implement the estimation algorithms of the item parameters in C language by using the item characteristics curve (ICC) for each item in the item response theory (IRT). As a conclusion of the study, experiments were tested with student’s exam scores, and showed difference errors between IRTC and BayesiAn program. Thus it showed the reliability of the system result by simulating a practical use possibility. The proposed c code can be thought as an alternative to be helpful for the development of additional programs on a web platform also. Further reseach is needed about the initial problem which is that eventually the error occurs depending on the initial value and the code implement problem on how to get an estimating parameters using other algorithms.

First-Order Context-Specific Likelihood Weighting in Hybrid Probabilistic Logic Programs

Statistical relational AI and probabilistic logic programming have so far mostly focused on discrete probabilistic models. The reasons for this is that one needs to provide constructs to succinctly model the independencies in such models, and also provide efficient inference.
 Three types of independencies are important to represent and exploit for scalable inference in hybrid models: conditional independencies elegantly modeled in Bayesian networks, context-specific independencies naturally represented by logical rules, and independencies amongst attributes of related objects in relational models succinctly expressed by combining rules.
 This paper introduces a hybrid probabilistic logic programming language, DC#, which integrates distributional clauses' syntax and semantics principles of Bayesian logic programs. It represents the three types of independencies qualitatively. More importantly, we also introduce the scalable inference algorithm FO-CS-LW for DC#. FO-CS-LW is a first-order extension of the context-specific likelihood weighting algorithm (CS-LW), a novel sampling method that exploits conditional independencies and context-specific independencies in ground models. The FO-CS-LW algorithm upgrades CS-LW with unification and combining rules to the first-order case.

DreamCoder: growing generalizable, interpretable knowledge with wake-sleep Bayesian program learning.

Expert problem-solving is driven by powerful languages for thinking about problems and their solutions. Acquiring expertise means learning these languages-systems of concepts, alongside the skills to use them. We present DreamCoder, a system that learns to solve problems by writing programs. It builds expertise by creating domain-specific programming languages for expressing domain concepts, together with neural networks to guide the search for programs within these languages. A 'wake-sleep' learning algorithm alternately extends the language with new symbolic abstractions and trains the neural network on imagined and replayed problems. DreamCoder solves both classic inductive programming tasks and creative tasks such as drawing pictures and building scenes. It rediscovers the basics of modern functional programming, vector algebra and classical physics, including Newton's and Coulomb's laws. Concepts are built compositionally from those learned earlier, yielding multilayered symbolic representations that are interpretable and transferrable to new tasks, while still growing scalably and flexibly with experience. This article is part of a discussion meeting issue 'Cognitive artificial intelligence'.

Comparison of the mathematical equation and trapezoidal approach for 24 h area under the plasma concentration-time curve calculation in patients who received intravenous vancomycin in an acute care setting.

The current recommendation for therapeutic monitoring of vancomycin has recently suggested AUC-guided dosing in patients with serious methicillin-resistant Staphylococcus aureus infections. The study objective was to evaluate mathematical equations and trapezoidal methods for calculating the 24 h area under the plasma vancomycin concentration-time curve (AUC24). The analysis of plasma vancomycin concentrations was performed in 20 adult patients treated with intravenous vancomycin. For each patient, AUC24 was estimated using two methods including, equation and trapezoidal calculation. The AUC24 from two methods was analyzed for correlation. The correlation between the equation and trapezoidal methods was strong. The coefficient of determination (R2 ) values was greater than .99. The two plasma vancomycin concentrations to achieve the highest correlation were concentration at 2.5 to 3h after starting the infusion and concentration at 1h before the next dose. Moreover, the AUC24 calculation from trapezoidal and equation methods showed that 19 out of 20 patients (95%) had AUC24 of more than 400 mg·h/L, and more than 50% in this group had AUC24/MIC greater than 600. Of those patients with AUC-trapezoidal >600, 15.38% of patients had trough under 15 mg/L, 15.38% of patients had trough in the range 15 to 20 mg/L and 69.23% of patients had trough more than 20 mg/L. The results of AUC-equation were similar to those of the AUC-trapezoidal method. Our study confirmed that the AUC monitoring is more appropriate than the trough vancomycin concentration. Given these considerations, the AUC-equation method is better and more practical to use in part of a point-of-care treatment, especially in the part of the Bayesian program is not available. The best sampling time point of the peak concentration was 0.5-1h after 2-h infusion.

941. Clinical Outcomes of Vancomycin Area-Under-the-Curve Monitoring: A Quasi-Experimental Study Interim Analysis

Abstract Background The 2020 Infectious Diseases Society of America vancomycin guidelines recommend the use of area-under-the-curve (AUC) monitoring instead of trough concentration monitoring. These guidelines changed due to evidence showing decreased rates of nephrotoxicity and lack of correlation between vancomycin troughs and patient treatment outcomes. Two common methodologies of calculating AUC exist in clinical practice, including two level kinetic calculations and Bayesian software programs. Previous studies used two levels for calculating the AUC. The purpose of this study is to evaluate if the same reduction in acute kidney injury (AKI) is achieved while using a Bayesian software program. Methods This retrospective quasi-experiment was performed at two sites within the same health system. The primary objective of this study was to assess the rate of AKI, as defined by the Acute Kidney Injury Network (AKIN) criteria, between patients receiving vancomycin dosed by trough monitoring as compared to AUC monitoring. Secondary objectives include mean daily dose of vancomycin, number of vancomycin levels, and severity of AKI. Patients were included in the study if they were 18 years or older, with at least one vancomycin concentration drawn. Patients were excluded if they had a baseline serum creatine greater than 2 mg/dL, received renal replacement therapy of any type, or received > 1 doses of vancomycin prior to admission. After screening for inclusion criteria, patients were matched 1:1 based on APACHE II score and vancomycin indication. Results One hundred and thirty patients met inclusion criteria at the interim analysis. Overall, baseline characteristics were similar between groups. No statistical difference was detected regarding the rate of acute kidney injury between the vancomycin trough group and AUC group (12% vs 14% p=0.971). The mean total daily doses of vancomycin were 1642 mg per day in the AUC group and 1838 mg per day in the trough group (p=0.08). Conclusion No statistical difference in rates of AKI were detected at the interim analysis, likely due to small sample size. Further data collection is ongoing to meet power and full results will be presented. Disclosures All Authors: No reported disclosures.

Bayesian method application: Integrating mathematical modeling into clinical pharmacy through vancomycin therapeutic monitoring.

The most recent consensus guidelines for dosing and monitoring vancomycin recommended the use of area-under-the-curve with Bayesian estimation for therapeutic monitoring. As this is a modern concept in the practice of clinical pharmacy, the main objective of this review is to introduce the fundamentals of Bayesian estimation and its mathematical application as it relates to vancomycin therapeutic drug monitoring. In addition, we aim to identify pharmacokinetic (PK) software programs that incorporate Bayesian estimation for vancomycin dosing and to describe the PK models utilized in those software programs for the adult population. Twelve software programs that utilize Bayesian estimation were identified, which included: Adult and Pediatric Kinetics, Best Dose, ClinCalc, DoseMeRx, ID-ODS, InsightRx, MwPharm++, NextDose, PrecisePK, TDMx, Tucuxi, and VancoCalc. The software programs varied in the population PK models used as the Bayesian a priori. With the presence of various vancomycin Bayesian software programs, it is important to choose those that utilize PK models reflective of the specific patient population.

Venetoclax Pharmacokinetics in Real World AML Patients: A Real Candidate for Therapeutic Drug Monitoring

Background Venetoclax (VNX), an orally bioavailable BCL-2 inhibitor, is now a gold standard treatment for newly diagnosed AML patients who were ineligible for intensive chemotherapy, in association with hypomethylating agent (HMA) or low-dose cytarabine (LDAC). Even if VNX pharmacokinetics (PK) have been well characterized and are described in the literature, only one clinical trials-based PK study has been conducted in AML patients. For example, bioavailability (F) and clearance (Cl) are largely influenced by food intake and CYP3A inhibitors (such as posaconazole PCZ), as shown by the 32% (F) and 44% (Cl) inter-individual variability for F and Cl, respectively. For such oral targeted therapies, real-world data highlighted a larger PK inter-variability (IIV) than expected. The most commonly reported toxicity under VNX treatment is neutropenia, frequently resulting in dose reduction and/or treatment discontinuation. In this study, we reported our local experience of VNX use with measure of PK exposure. Material and Methods We retrospectively evaluated patients treated by VNX + azacitidine (AZA) for AML with available PK blood sample. Briefly, VNX was administered at a dose of 400mg (or 70mg if co-medication with PCZ) once daily, and PK was mainly performed from C1D4 to C1D7 (T0, T4, T8) during hospitalization. In order to estimate individual PK parameters in our cohort, we integrated a published PK population (PKpop) model2 into the Bayesian software program Tucuxi®. For each patient, model was used to fit the VNX dataset, which basically computes the individual Bayesian posterior parameter values based on the prior information. Goodness-of-fit was assessed by computing bias and imprecision of individual model predicted concentrations. Then, individual area under the concentration-time curves (AUC) and PK parameters were computed. In parallel, the following outcomes have been collected in order to evaluate any potential PK-PD relationships. All informed consents have been obtained according to the reference methodology MR-004. Results Twenty patients (H/F = 12/8, mean age = 66 years [min 38 max 77]) were included in this study, they all received in average 4 courses of VNX-AZA (Table 1). After a median follow-up of 152 days [27-642], 65% of patients were in CR + CRi. All patients who did not achieve CR, received VNX-AZA in second line after failure of previous chemotherapy. Minimal residual disease (MRD) assessed by flow cytometry was negative in 40% of patients. At the last follow up, 11 patients (55%) were alive in CR and 1 patient (5%) in relapse. Eight patients (40%) died from progression in 6 patients and toxic death in 2 patients. All patients experienced at least one hematological toxicity grade ≥ 3. Between November 2021 and June 2022, 93 PK blood sampling were analyzed (80% in course 1). Nine patients had a second day of supplementary PK. Seven patients received concomitant PZC and 3 did not take VNX with food. PKpop model applied to our cohort showed a satisfactory goodness-of-fit (Figure 1, R²=0.83, bias of 2.8% and imprecision of 26.5%). Mean individual daily AUC was 35.8 mg.h/L with high IIV (CV=47%) and no difference according to PCZ use or food intake (p=NS). No PK-PD relationships (efficacy and toxicity) has been highlighted. Conclusion This work described exposure and PK variability in real-world setting. Even if some covariates have been highlighted as sources of variability, it seems difficult to control exposure with a "one size fits all" dosing. PK variability could be an explanation for PD variability (both efficacy and toxicity). VNX seems to be a good candidate for therapeutic drug monitoring but some PK-PD relationships studies need to be conducted to find precise PK target. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal

Critical assessment of the revised guidelines for vancomycin therapeutic drug monitoring

BackgroundThe revised vancomycin guidelines recommend replacing trough-only with trough or peak/trough Bayesian and first-order equations monitoring, citing their better AUC predictions and poor AUC-trough R2. Yet, evidence suggesting good AUC-trough correlation has been overlooked, and the optimality of peak/trough samples has been doubted. The guidelines recommend Bayesian programs implement richly-sampled PopPK priors despite their scarcity. Therefore, whether complex Bayesian and sample-demanding first-order equations can bring significant advantages to the practice over simple trough-only monitoring is worth weighing. ObjectivesThe primary aim is to compare the predictive performance of the AUC monitoring methods. Then, we investigate the impact of not adhering to trough sampling on the Bayesian-based predictions. Moreover, we report the nature of PopPK priors used in Bayesian programs to assess the applicability of the guideline recommendations. MethodsWe calculated the predictive performance of the monitoring methods using a standard PopPK modeling and simulation approach. We thoroughly explored the prior PK models implemented in Bayesian programs. ResultsPredictive performances of the monitoring methods were comparable at steady-state relative to the number of samples. Contrary to the recommendation, Bayesian trough monitoring did not result in better predictive performances compared to using random levels. Very few programs implemented richly-sampled priors. ConclusionAll the monitoring methods can be, relatively, reliable at steady-state, if properly implemented. Although only Bayesian-based monitoring can be used pre-steady-state, its predictive performance can be modest. Trough-only monitoring is the simplest approach. Constraints regarding trough sampling times could be relaxed. The scarcity of richly-sampled Bayesian priors questions the applicability of the revised guidelines recommendation.

Synthesizing theories of human language with Bayesian program induction

Automated, data-driven construction and evaluation of scientific models and theories is a long-standing challenge in artificial intelligence. We present a framework for algorithmically synthesizing models of a basic part of human language: morpho-phonology, the system that builds word forms from sounds. We integrate Bayesian inference with program synthesis and representations inspired by linguistic theory and cognitive models of learning and discovery. Across 70 datasets from 58 diverse languages, our system synthesizes human-interpretable models for core aspects of each language’s morpho-phonology, sometimes approaching models posited by human linguists. Joint inference across all 70 data sets automatically synthesizes a meta-model encoding interpretable cross-language typological tendencies. Finally, the same algorithm captures few-shot learning dynamics, acquiring new morphophonological rules from just one or a few examples. These results suggest routes to more powerful machine-enabled discovery of interpretable models in linguistics and other scientific domains.

Seismic Hazard Evaluation of the North and Northwest Iranian Plate Considering the Bayesian Perspective

In the last hundred years alone, there have occurred more than 2000 earthquakes larger than 4 in the northern and northwestern parts of Iran. Because Iran is placed in the heart of the Alpine-Himalayan collision zone, which is one of the world’s greatest seismic locations, massive earthquakes have regularly struck the country. Seismic hazard assessment is a realistic and helpful method for engineering seismology to prevent motions in a location in order to forecast the occurrence of earthquakes. For parameter estimation in seismic hazard assessments, the Bayesian probability theory provides a practical and accurate technique. As a result, Bayesian techniques for analysis are based on an older method. Furthermore, they have the unique capacity to account for the variability of characteristics in probabilistic relationships. In this assessment, we used the Bayesian earthquake risk estimation program developed by Lyubushin. Our research is in the sequence of applications of this program to assess the risk of earthquakes in various parts of the universe. In this method, the basic assumption is that earthquakes follow the characteristic of being Poisson. Its zoning map is drawn in the north and northwest regions of Iran. The highest value was obtained for the city of Tabriz with 0.29 galls, and the lowest value was obtained for the city of Zanjan with the value of 0.06 galls.