Multistage Sequence Research Articles

Code-Aware Prompting: A Study of Coverage-Guided Test Generation in Regression Setting using LLM

Testing plays a pivotal role in ensuring software quality, yet conventional Search Based Software Testing (SBST) methods often struggle with complex software units, achieving suboptimal test coverage. Recent work using large language models (LLMs) for test generation have focused on improving generation quality through optimizing the test generation context and correcting errors in model outputs, but use fixed prompting strategies that prompt the model to generate tests without additional guidance. As a result LLM-generated testsuites still suffer from low coverage. In this paper, we present SymPrompt, a code-aware prompting strategy for LLMs in test generation. SymPrompt’s approach is based on recent work that demonstrates LLMs can solve more complex logical problems when prompted to reason about the problem in a multi-step fashion. We apply this methodology to test generation by deconstructing the testsuite generation process into a multi-stage sequence, each of which is driven by a specific prompt aligned with the execution paths of the method under test, and exposing relevant type and dependency focal context to the model. Our approach enables pretrained LLMs to generate more complete test cases without any additional training. We implement SymPrompt using the TreeSitter parsing framework and evaluate on a benchmark challenging methods from open source Python projects. SymPrompt enhances correct test generations by a factor of 5 and bolsters relative coverage by 26% for CodeGen2. Notably, when applied to GPT-4, SymPrompt improves coverage by over 2x compared to baseline prompting strategies.

MMSSC-Net: multi-stage sequence cognitive networks for drug molecule recognition

MMSSC-Net: multi-stage sequence cognitive networks for drug molecule recognition

A multi-agent double Deep-Q-network based on state machine and event stream for flexible job shop scheduling problem

With the industry's rapid shift to large-scale personalized development, the need for a ramose multi-agent system capable of flexible job shop scheduling problem (FJSP) is apparent. As such, this paper proposes a novel framework that transforms the combinatorial optimization problem into a multi-stage sequence decision problem and introduces a multi-agent double Deep-Q-network algorithm (MADDQN) for FJSP. The event-driven workshop environment model based on state machine and event stream mechanism is constructed to transform FJSP into a Markov Decision Process (MDP) with more robust general performance and decouple the workshop environment model from the decision analysis model. The developed multi-agent reinforcement learning consisting of job agents and machine agents applies dispatching rules to maximize the cumulative reward of all agents by Boltzmann exploitation to avoid the local optimum. Results of numerical experiments reflect that the proposed MADDQN performs better than the traditional approaches in large-scale instances, and the real-time performance of the scheduling decision is guaranteed.

Adaptive Probabilistic Optimization Approach for Vibration-Based Structural Health Monitoring

This paper presents a novel adaptive probabilistic algorithm to identify damage characteristics by integrating the use of the frequency response function with an optimization approach. The proposed algorithm evaluates the probability of damage existence and determines salient details such as damage location and damage severity in a probabilistic manner. A multistage sequence is used to determine the probability of damage parameters including crack depth and crack location while minimizing uncertainties. A simply supported beam with an open edge crack was used to demonstrate the application of the algorithm for damage detection. The robustness of the algorithm was tested by incorporating varying levels of noise into the frequency response. All simulation results show successful detection of damage with a relatively high probability even in the presence of noise. Results indicate that the probabilistic algorithm could have significant advantages over conventional deterministic methods since it has the ability to avoid yielding false negatives that are quite common among deterministic damage detection techniques.

JolasMATIKA: An Experience for Teaching and Learning Computing Topics From University to Primary Education

Contribution: A learning-by-teaching methodology through games can be used to promote informatics (computer science) in primary and secondary education. Applying the proposed activities can change students' perception of informatics from seeing it as merely using computers to seeing its relationship with mathematics. The experience can also help students acquire competences in teaching. Background: Although students, specifically in primary and secondary education, are increasingly competent in terms of technology use, it has been found that in many cases informatics, as a science, has been relegated to a secondary status; it is usually considered only as a tool or additional resource, and not as an object of study. Intended Outcomes: To refine the application of the learning-to-teach-to-learn (L2T2L) methodology, a learning-by-teaching methodology that has students learn and then, in turn, teach that learning to younger students, in cascade from university to secondary to primary students. To analyze its effects on students' attitudes toward informatics. Application Design: The model incorporates a learning-by-teaching approach in a multistage sequence across different kinds of learners and teachers, using fun, game-like materials. Findings: The use of the action research methodology allowed adjustment of the educational methodology, providing more reliable data and enough experience to suggest how to extend the project to a broader audience. Although the results obtained were less significant than expected, the experience did give students a more realistic view of informatics.

Multistage Axial Hydro-Forging Sequence: A New Forming Approach for Manufacturing of Double-Stepped Tubes

To manufacture the large-expansion-ratio double-stepped tube without thinning, a new approach known as multistage axial hydro-forging sequence has been proposed in this study. The whole forming process is divided into five stages. Compared with traditional tube hydroforming, the thickness can be increased by axial compression deformation to reduce the thinning ratio and enhance the forming capacity of the tube in hydro-forging stage and double-stepped hydro-forging stage, respectively. The proposed multistage axial hydro-forging sequence process was implemented by experiments along with the finite element simulation to validate its feasibility. The suitable loading path to prevent defects such as wrinkle and thinning was discussed. The results showed that double-stepped tube with large expansion ratio can be produced successfully with movable dies’ design. There is no thinning in the final tube, particularly in all stepped area, and a better thickness distribution could be achieved. As a result, the proposed multistage axial hydro-forging sequence can be a feasible forming approach for the fabrication of large-expansion-ratio double-stepped tubes with high reliability.

A commercial laccase-mediator system to delignify and improve saccharification of the fast-growingPaulownia fortunei(Seem.) Hemsl.

AbstractIt was demonstrated for the first time that a laccase-based enzymatic pretreatment is able to delignify fast-growing paulownia species. The treatment was performed with a commercial low-redox potential laccase isolated fromMyceliophthora thermophila(Apinis) Oorschot and methyl syringate (MeS) as a natural phenolic mediator. Up to 24% lignin removal was attained by the laccase-MeS treatment (L/MeS), followed by alkaline peroxide extraction in a multistage sequence. The reduction in lignin content was accompanied by a significant improvement in the subsequent enzymatic saccharification, with increases of up to 38% glucose and 34% xylose yields. The structural modifications of the lignin were analyzedin situby two dimensional-nuclear magnetic resonance (2D-NMR) spectroscopy. A considerable removal of guaiacyl and syringyl lignin units with respect to the carbohydrate signals was visible as well as the cleavage of β-O-4′, β-5′ and β-β′ linkages leading to elevated amounts of Cα-oxidized guaiacyl and syringyl units. The presence of oxidized lignin compounds in the filtrates of the enzymatic treatments – such as vanillin, vanillic acid, syringaldehyde and syringic acid – conclusively demonstrates the ability of L/MeS treatment to oxidize and depolymerize the lignin in paulownia wood.

Structuring Surfaces by Microfinishing Using Defined Abrasive Belts

Microfinishing, also known as superfinishing or short-stroke honing, is a commonly used process for generating technical surfaces focusing on tribological applications. Due to microfinishing processes high surface qualities are manufacturable regarding the surface roughness and bearing area ratio. While the required characteristics for tribological loaded workpieces are changing with their rising significance, the surface structuring is becoming more and more important. With the use of defined abrasive belts, the possibilities of surface structuring by microfinishing are enhanced. The possibilities and challenges concerning surface structuring by microfinishing applying defined abrasive belts are described in this research study. Therefore, a geometrical-kinematic simulation is used to predict the theoretical structures generated by microfinishing, while in experimental investigations the influences of kinematic parameters and a multi-stage process sequence are considered.

Behavior of pile group with elevated cap subjected to cyclic lateral loads

The pile group with elevated cap is widely used as foundation of offshore structures such as turbines, power transmission towers and bridge piers, and understanding its behavior under cyclic lateral loads induced by waves, tide water and winds, is of great importance to designing. A large-scale model test on 3×3 pile group with elevated cap subjected to cyclic lateral loads was performed in saturated silts. The preparation and implementation of the test is presented. Steel pipes with the outer diameter of 114 mm, thickness of 4.5 mm, and length of 6 m were employed as model piles. The pile group was cyclic loaded in a multi-stage sequence with the lateral displacement controlled. In addition, a single pile test was also conducted at the same site for comparison. The displacement of the pile cap, the internal forces of individual piles, and the horizontal stiffness of the pile group are presented and discussed in detail. The results indicate that the lateral cyclic loads have a greater impact on pile group than that on a single pile, and give rise to the significant plastic strain in the soil around piles. The lateral loads carried by each row of piles within the group would be redistributed with loading cycles. The lateral stiffness of the pile group decreases gradually with cycles and broadly presents three different degradation patterns in the test. Significant axial forces were measured out in some piles within the group, owing to the strong restraint provided by the cap, and finally lead to a large settlement of the pile group. These findings can be referred for foundation designing of offshore structures.

In-Depth 2D NMR Study of Lignin Modification During Pretreatment of Eucalyptus Wood with Laccase and Mediators

Eucalypt feedstock was pretreated with Myceliophthora thermophila laccase and methyl syringate, as mediator, in a multistage sequence consisting of successive enzymatic and alkaline peroxide stages, directly on the ground wood. Lignin modification was studied by two-dimensional (2D) nuclear magnetic resonance (NMR) of wood (at the gel state) after each stage of the sequence. Decrease of guaiacyl and syringyl lignin units (with preferential removal of the former) and aliphatic (mainly β-O-4′-linked) side chains was observed along the sequence, without a substantial change in polysaccharide cross-signals along the sequence, was observed. However, the most noticeable modification was the formation of Cα-oxidized syringyl units in the enzymatic stages, which were partially removed in the alkaline extractions. Further insight was attained by 2D NMR of the cellulolytic-enzyme lignins, isolated from the pretreated samples, and the filtrates from the different stages. Additionally, a comparison of eucalypt delignification with the high redox potential laccase from Pycnoporus cinnabarinus in the presence of the synthetic mediator 1-hydroxybenzotriazole was carried out. Wood delignification already occurred at the first cycle of the pretreatment with P. cinnabarinus laccase and 1-hydroxybenzotriazole, which directly correlated with increases in glucose (~10 %) yields after enzymatic hydrolysis. In contrast, the first cycle of pretreatment with M. thermophila laccase and methyl syringate did not produce any delignification or saccharification improvement. However, at the end of the sequence, similar delignification (~50 %) of eucalypt feedstock and increases (~30 %) in glucose yields were attained in both laccase-mediator pretreatments. Finally, the influence of lignin content and the ratio of its syringyl and guaiacyl units on recalcitrance of eucalypt feedstock to enzymatic saccharification is discussed.

Pretreatment with laccase and a phenolic mediator degrades lignin and enhances saccharification of Eucalyptus feedstock.

BackgroundBiofuel production from lignocellulosic material is hampered by biomass recalcitrance towards enzymatic hydrolysis due to the compact architecture of the plant cell wall and the presence of lignin. The purpose of this work is to study the ability of an industrially available laccase-mediator system to modify and remove lignin during pretreatment of wood (Eucalyptus globulus) feedstock, thus improving saccharification, and to analyze the chemical modifications produced in the whole material and especially in the recalcitrant lignin moiety.ResultsUp to 50% lignin removal from ground eucalypt wood was attained by pretreatment with recombinant Myceliophthora thermophila laccase and methyl syringate as mediator, followed by alkaline peroxide extraction in a multistage sequence. The lignin removal directly correlated with increases (approximately 40%) in glucose and xylose yields after enzymatic hydrolysis. The pretreatment using laccase alone (without mediator) removed up to 20% of lignin from eucalypt wood. Pyrolysis-gas chromatography/mass spectrometry of the pretreated wood revealed modifications of the lignin polymer, as shown by lignin markers with shortened side chains and increased syringyl-to-guaiacyl ratio. Additional information on the chemical modifications produced was obtained by two-dimensional nuclear magnetic resonance of the whole wood swollen in dimethylsulfoxide-d6. The spectra obtained revealed the removal of guaiacyl and syringyl lignin units, although with a preferential removal of the former, and the lower number of aliphatic side-chains per phenylpropane unit (involved in main β-O-4ʹ and β-βʹ inter-unit linkages), in agreement with the pyrolysis-gas chromatography/mass spectrometry results, without a substantial change in the wood polysaccharide signals. However, the most noticeable modification observed in the spectra was the formation of Cα-oxidized syringyl lignin units during the enzymatic treatment. Further insight into the modifications of lignin structure, affecting other inter-unit linkages and oxidized structures, was attained by nuclear magnetic resonance of the lignins isolated from the eucalypt feedstock after the enzymatic pretreatments.ConclusionsThis work shows the potential of an oxidative enzymatic pretreatment to delignify and improve cellulase saccharification of a hardwood feedstock (eucalypt wood) when applied directly on the ground lignocellulosic material, and reveals the main chemical changes in the pretreated material, and its recalcitrant lignin moiety, behind the above results.

Open Die Forging of Large Shafts with Defects – Physical and Numerical Modelling

The aim and scope of this paper is centered to analyze the influence of the geometry of V-shaped dies on the closure of internal centerline porosity defects in ingots during multistep open-die forging. The investigation is performed with small scale physical models made from lead using V-shaped dies with 90º and 120º and a reference pair of flat parallel platens. Holes drilled through the center of these preforms are produced to mimic centerline porosity in full scale cast ingots and intermediate rotation of the preforms replicate a multi-stage forging sequence under laboratory testing conditions. The presentation is supported by finite element modelling using an in-house developed computer program and the overall investigation shows that better results in closure of centerline defects are obtained with a V-shaped die with 120º die angle.

Structure of Travel Planning Processes and Information Use Patterns

Based on an empirical data set collected from Chinese tourists visiting Macau SAR, this study attempts to decompose the travel decision-making process into a multistage sequence (before purchase, at time of purchase, after purchase, after arriving at destination) with multiple decision choices for a vacation trip. The study results demonstrate that travel decision making follows a multidimensional, ongoing sequence and is a hierarchical process. This study further attempts to expand the framework for describing the travel decision-making structure by incorporating information use patterns across the overall course of the travel planning process. The contingent nature of decision making is partially validated by tracing the sequence of decision making and information use along the travel planning horizon. Furthermore, given the lack of studies based on Chinese tourists’ decision-making behavior, the study results provide new insights for many tourism destinations and businesses eyeing the burgeoning Chinese tourists’ market.

Defining inventory control points in multiproduct stochastic pull systems

Multistage pull production systems have been widely implemented in recent years and constitute a significant aspect of lean manufacturing. One of the important considerations in such systems is identifying the control points, i.e. where in the multistage sequence to locate the output buffers. Allowable container/batch sizes, optimal inventory levels, and ability of systems to automatically adjust to stochastic demand depend on the location of these control points yet the issue of optimal location has not been widely addressed. This paper considers a multiproduct pull setting where part types compete with each other for common production resources. In this environment it is important to consider factors such as lead time variability and to include the corresponding queuing aspects into the model. Each workstation is modeled as a GI/G/1 queue. Waiting times spent by parts at workstations are approximated using a decomposition/recomposition algorithm. Necessary and sufficient conditions are provided for the optimality of a single control point. Conditions under which multiple control points are optimal are investigated along with the impact of product mix and utilization parameters on the number of control points. Analytical model results are validated by simulation.

Natural products and colon cancer: current status and future prospects

Carcinogenesis is a multistage process consisting of initiation, promotion and progression phases. Thus, the multistage sequence of events has many phases for prevention and intervention. Chemoprevention, a novel approach for controlling cancer, involves the use of specific natural products or synthetic chemical agents to reverse, suppress or prevent premalignancy before the development of invasive cancer. Several natural products, such as, grains, nuts, cereals, spices, fruits, vegetables, beverages, medicinal plants and herbs and their various phytochemical constituents including, phenolics, flavonoids, carotenoids, alkaloids, nitrogen containing as well as organosulfur compounds confer protective effects against wide range of cancers including colon cancer. Since diet has an important role in the etiology of colon cancer, dietary chemoprevention received attention for colon cancer prevention. However, identification of an agent with chemopreventive potential requires in vitro studies, efficacy and toxicity studies in animal models before embarking on human clinical trials. A brief introduction about colon cancer and the role of some recent natural products in colon cancer chemoprevention with respect to multiple molecular mechanisms in various in vitro, in vivo and clinical studies are described in this review.

SELECTION OF INVENTORY CONTROL POINTS IN MULTISTAGE PULL SYSTEMS

This paper considers a multistage, serial production system using pull control for discrete parts manufacturing. A model is developed to locate inventory control points, i.e. where in the multistage sequence to locate the output buffers. Allowable container/batch sizes, optimal inventory levels, and ability of systems to automatically adjust to stochastic demand depend on the location of these control points. Necessary and sufficient conditions are derived for ensuring that a single, end-of-line accumulation point is optimal. When this is not the case, an algorithm is provided to determine the optimal control points.

Formation of elliptical garnet in a metapelitic enclave by melt‐assisted dissolution and reprecipitation

AbstractMetapelitic residual enclaves in the Neogene Volcanic Province of SE Spain are residues left after melt extraction. Glass (quenched melt) of granitic composition occurs as inclusions in most minerals and as intergranular pockets. The most common enclave types show one stage of garnet growth that is interpreted to have occurred at the same time as glass production. Some of these show a well‐developed foliation outlined by fibrolite, biotite, graphite and glass, which wraps around elongate garnet crystals that have aspect ratios up to 10:1. Based on microstructures and chemistry, the garnet within these rocks shows clear core and mantle structure. The core has an average composition of Alm76–Prp08–Sps14–Grs03 and contains primary inclusions of biotite and melt, trapped during garnet growth. A thin (c. 100 μm), irregular mantle overgrows the garnet core, enclosing oriented fibrolite inclusions in strain caps, and biotite in strain shadows. In places, the overgrowths form skeletal elongated arms, which extend parallel to the foliation. The garnet mantle contains less Mn and higher XMg, but both core and mantle display flat Mn profiles, the contact being a sharp break. Ternary feldspar and Grt–Bt thermometry yield temperatures in the range 800–900 °C, with no systematic differences among the different microstructural domains of elliptical garnet. Based on the observed intracrystalline microstructures, the high amount of melt extraction in the rock by flattening component strain and the chemical zoning of garnet, the formation of elliptical garnet is modelled by a multistage sequence. This involves pressure solution and reprecipitation of the core, followed by post‐kinematic, partly mimetic growth of the garnet mantle.

The Class Action as a Financial Call Option

Class action law enforcement is a resource-intensive undertaking, requiring investment of both intellectual capital and financial wherewithal. Building on insights developed in the economic theory of investment under uncertainty, the key theoretical proposition is that class action law enforcement comprises a multi-stage sequence of investment opportunities under conditions of multi-dimensional uncertainty. The properties of investment in class actions include (i) investment expenditures; (ii) future rewards on investment; (iii) multi-dimensional investment uncertainty; (iv) the irreversibility of investment expenditures; and (v) the sequential, multi-stage property of investment opportunities and investment decisions. The court appointment of class counsel is modeled as a judicially-granted monopoly over investment in class action law enforcement. At any given stage of the sequence (but for the last one), plaintiffs' attorneys and, later, court-appointed class counsel are faced with a financial call option, namely, an opportunity to invest and buy, with some probability, the opportunity to invest in the stage that follows in the sequence. Plaintiffs' attorneys' and, subsequently, the court-appointed class counsel's incentives to invest and investment decisions throughout the sequence are overwhelmingly the most important determinant of the magnitude of liability exposure. These theoretical propositions provide the most conceptually-inclusive model of the incentive structure of class action law enforcement. This analytic model is conceptually-inclusive because the myriad decisions made by plaintiffs' attorneys and, subsequently, by class counsel in the course of class action law enforcement are perceived as sequential investment decisions under conditions of uncertainty. Thus, virtually any aspect relating to the conduct and performance of plaintiffs' attorneys becomes amenable to analysis as a sequential investment problem. Normative implications concerning the design of investment-oriented, welfare-enhancing regulation of class action law enforcement across different areas of law where class actions are used as an enforcement mechanism are derived.

The Law Enforcement Venture: Understanding the Effects of Investment in Class Actions on Corporate Liability Exposure

Class action law enforcement involves a multi-stage sequence of options to invest (similar to financial call options) under conditions of uncertainty. This general theoretical proposition sets forth a conceptually inclusive model of the underlying incentive structure of class action law enforcement across virtually any area of economic and social activity in which the class action mechanism is used to enforce the law. Under this model, the myriad aspects of plaintiffs' attorneys' effort and performance boil down to sequential options to invest and sequential investment decisions. Against this backdrop, this Article examines the compound correlation between (i) the magnitude of plaintiffs' attorneys' incentives to invest; (ii) the level of investment made in the various stages of the class action sequence; (iii) the magnitude of corporate liability exposure for harm caused through systematic wrongful conduct; and (iv) the deterrent and remedial effects of class action law enforcement. Focusing on corporate liability exposure, the inquiry specifically examines the effect of investment on (i) the probability of detecting systematic wrongful behaviors; (ii) the probability of holding corporate defendants liable for the social cost of systematic wrongdoing; (iii) the magnitude of average damages awarded to class members compared with the average harm actually caused; and (iv) the size of the plaintiffs' class compared with the size of the victims' population. The analysis shows how the magnitude of plaintiffs' attorneys' incentives to invest and their corresponding investment decisions interact with other factors to determine the magnitude of liability exposure. While this correlation is sensitive to the economic attributes of the plaintiffs' law firm and the law enforcement context, the analysis generates unambiguous insights concerning its general properties. These insights are germane to identifying the existence of a discrete, privately- and socially-optimal level of investment in class action law enforcement. These insights also provide the intellectual foundations necessary for crafting welfare-enhancing social policy and designing effective regulatory measures that bridge the private-social divergence of incentives to invest in class action law enforcement. Possible investment-oriented regulatory measures are identified and discussed.

The Numerical Simulation of an Unbalanced Jetlet and Its Role in the Palm Sunday 1994 Tornado Outbreak in Alabama and Georgia

Abstract Meso-beta-scale numerical model simulations and observational data are synthesized in an effort to develop a multistage paradigm for use in forecasting tornadic convection in the southeastern United States. The case study to be utilized as an example of the multistage sequence of events is the Palm Sunday 1994 outbreak, which culminates with the development of an unbalanced mesoscale jet streak or jetlet that focuses a given region for significant values of low-level vertical wind shear, low-level confluence and vertical vorticity, midtropospheric cooling, and storm-relative helicity. The five-stage paradigm includes 1) the existence of a jet exit region accompanying a deep balanced thermally indirect circulation south of the outbreak and a return branch ageostrophic low-level southerly jet, both typically accompanying the subtropical jet stream and the leading edge of hot continental air; 2) the existence of a jet entrance region accompanying a deep balanced thermally direct circulation north of...