Speed-quality Trade-off Research Articles

High-Quality Hypergraph Partitioning

Hypergraphs are a generalization of graphs where edges (aka nets ) are allowed to connect more than two vertices. They have a similarly wide range of applications as graphs. This article considers the fundamental and intensively studied problem of balanced hypergraph partitioning (BHP) , which asks for partitioning the vertices into k disjoint blocks of bounded size while minimizing an objective function over the hyperedges. Here, we consider the two most commonly used objectives: the cut-net metric and the connectivity metric . We describe our open-source hypergraph partitioner KaHyPar which is based on the successful multi-level approach—driving it to the extreme of using one level for (almost) every vertex. Using carefully designed data structures and dynamic update techniques, this approach turns out to have a very good time–quality tradeoff. We present two preprocessing techniques— pin sparsification using locality-sensitive hashing (LSH) and community detection based on the Louvain algorithm . The community structure is used to guide the coarsening process that incrementally contracts vertices. Portfolio-based partitioning of the contracted hypergraph then already achieves a good initial solution. While reversing the contraction process, a combination of several refinement techniques achieves a good final partitioning. In particular, we support a highly-localized local search that can directly produce a k -way partitioning and complement this with flow-based techniques that take a more global view. Optionally, a memetic algorithm evolves a pool of solution candidates to an overall good solution. We evaluate KaHyPar for a large set of instances from a wide range of application domains. With respect to quality, KaHyPar outperforms all previously considered systems that can handle large hypergraphs such as hMETIS, PaToH, Mondriaan, or Zoltan. Somewhat surprisingly, to some extend, this even extends to graph partitioners such as KaHIP when considering the special case of graphs. KaHyPar is also faster than most of these systems except for PaToH which represents a different speed–quality tradeoff.

LIDER

Passage retrieval has been studied for decades, and many recent approaches of passage retrieval are using dense embeddings generated from deep neural models, called "dense passage retrieval". The state-of-the-art end-to-end dense passage retrieval systems normally deploy a deep neural model followed by an approximate nearest neighbor (ANN) search module. The model generates embeddings of the corpus and queries, which are then indexed and searched by the high-performance ANN module. With the increasing data scale, the ANN module unavoidably becomes the bottleneck on efficiency. An alternative is the learned index, which achieves significantly high search efficiency by learning the data distribution and predicting the target data location. But most of the existing learned indexes are designed for low dimensional data, which are not suitable for dense passage retrieval with high-dimensional dense embeddings. In this paper, we propose LIDER , an efficient high-dimensional L earned I ndex for large-scale DE nse passage R etrieval. LIDER has a clustering-based hierarchical architecture formed by two layers of core models. As the basic unit of LIDER to index and search data, a core model includes an adapted recursive model index (RMI) and a dimension reduction component which consists of an extended SortingKeys-LSH (SK-LSH) and a key re-scaling module. The dimension reduction component reduces the high-dimensional dense embeddings into one-dimensional keys and sorts them in a specific order, which are then used by the RMI to make fast prediction. Experiments show that LIDER has a higher search speed with high retrieval quality comparing to the state-of-the-art ANN indexes on passage retrieval tasks, e.g., on large-scale data it achieves 1.2x search speed and significantly higher retrieval quality than the fastest baseline in our evaluation. Furthermore, LIDER has a better capability of speed-quality trade-off.

Fine motor skills and motor control networking in developmental age.

ObjectivesWe investigated the relationships between fine motor skills, fitness, anthropometrics, gender and perceived motor performance in school beginners. The aim of our study was to delineate whether and to what extent fine motor control would show meaningful synchrony with other motor variables in the age of onset of handwriting in school.MethodsA sample of N = 239 of 6‐to‐8‐year‐old children were tested with an array of tasks measuring fine motor (i.e., dexterity and speed) and grapho‐motor performance (tracing on a tablet screen), anthropometric indexes, and fitness (shuttle run) measures. A subset of 95 children was also tested for perceived motor competence.ResultsIn spite of an overall poor anthropometric condition, our participants were relatively fit. As expected, older children performed better in both, fine motor tasks and the shuttle test. The girls were better in fine motor skills, and an original speed‐quality trade‐off in the drawing was found. However, the magnitude of difference by grade was greater for boys' fine motor skills than those of girls'. A network analysis revealed three specific clusters, (1) perceived competencies, (2) fitness, and (3) fine motor skills.ConclusionsGiven the relative independence of these areas of physical performance, we suggest focusing on these three clusters as distinct areas of physical education. Fine motor skills deserve further consideration, especially at an early school age. We have demonstrated that network analysis and technology devices used to evaluate motor development are useful and meaningful tools.

Does regulatory approval speed impact product quality? Evidence from the US pharmaceutical industry (WITHDRAWN)

The trade-off between speed to market and product quality is a well-studied phenomenon in the new product development literature. However, there is little evidence of how supply chain stakeholders beyond the focal firm can impact the speed-quality trade-off. Moreover, the impact of speed to market on performance parameters outside the firm – such as consumer safety and surplus, or social welfare – is unclear. We address both these gaps in the literature by studying speed to market from a regulatory perspective and analyzing the impact of speed to market on safety-related construct. Rooted in the context of the US pharmaceutical drug industry, our study evaluates the impact of four expedited drug approval programs by the USFDA on the quality of the newly launched pharmaceutical drugs in the US. These programs are designed to provide speedier access to those novel drugs which address a large unmet clinical need for the patients. We operationalize product quality as the number of adverse events and show that drugs approved under expedited programs could result in as high as 131% more number of adverse events and 167% more number of serious adverse events. Our study makes important theoretical and practical contributions. Theoretically, this is the first study in the NPD literature that, to the best of our knowledge, examines the impact of regulatory speed on the product quality. Practically, our study highlights critical unintended consequences of expediting the drug approval process for the regulators.

Managing Discretionary Services: A Review and Research Opportunities

Discretionary services typically refer to professional work and complex service work by physicians, software developers, web designers, lawyers, or financial analysts, where there are no standard working processes and customer perceived quality of service increases with the time spent on it. Recently, research on these services, especially the corresponding speed-quality tradeoff problem, has gained more and more attention. This paper reviews both the analytical models and the empirical studies in this area, highlighting their contributions and pointing out potential directions for future research.

Using queue to signal quality in primary care: The need for further empirical investigation.

In recent times, signalling quality via queues for generic services has received significant attention. However, the literature till date on primary care services has focussed on the optimal speed-quality trade-off to ensure higher service times and lower waiting times for the patients. Borrowing from the queue management literature on generic services, we aim to understand whether the hypothesis that queue is a signal of physician's quality is reasonable. Based on theoretical arguments, we justify the need to investigate this hypothesis on queue acting as a signal of quality in primary care. Although we conjecture that the queue length on arrival may act as a signal of quality, the waiting experienced after joining the queue is still considered expensive.

Speed-Quality Tradeoffs in Home Health: The Effects of Visit Length on Hospital Readmission

Home health care is a rapidly growing area of the health sector in the United States. We study its role in the shift towards value-based care, as it is viewed as an avenue for achieving reductions in the cost and utilization of expensive downstream health care services. Using a novel dataset on home health care visits, we examine whether and how the amount of time that a provider spends during a home health visit with a recently-discharged patient impacts the patient’s likelihood of being readmitted to the hospital. Since unobserved patient health status may influence both the length of a home health visit and the likelihood of hospital readmission, we use the within-provider average visit length of all other episodes’ visits conducted by each provider in the 30-day period before and after the focal visit as an instrument for visit length. Using this instrumental variable approach and controlling for operational, demographic, and patient condition-related characteristics, we find the following: on average, an extra minute during a focal home health visit is associated with a 1.39% decrease in the likelihood of readmission to the hospital following that visit. Our finding suggests that a 10% increase in visit length would decrease the likelihood of readmission following a home health visit by 6%. We document heterogeneity in this effect across different patient types and visit types. We conduct a cost-benefit analysis that suggests that the cost of investing in additional home health capacity is outweighed by the cost savings arising from fewer hospitalizations.

Selling Discretionary Services to Strategic Customers with Peripheral Consumption

Discretionary service industries are those in which customers prefer not to wait in line, but also derive more value from a longer service time; this is common in healthcare interactions or repair settings. In such settings, the service provider often can obtain not only a fixed fee for the service but also a proportion of the price the customer pays for tests and materials consumed in the service process. As the prices customers pay for such peripherals (e.g. medical imaging and drugs in healthcare, spare parts in repairs) are often externally determined (by insurance, government, or third-party providers), their contribution to the service provider's income complicates the typical speed-quality trade-off decision. We investigate this setting, revealing new insights based on a strategic queueing framework in which the service provider decides the service fee and service rate, an exogenous entity determines the price the customers pay for peripherals, and the customers choose whether to patronize the service by comparing the service value and total costs. We find that if the price of the peripherals increases, the service provider reduces the service fee and increases the service time to improve the service quality, but the equilibrium demand rate still deceases. This effect is more profound if the service provider's share of the peripheral revenues is smaller. We also find that in the presence of peripheral consumption, when the customers become more sensitive to service quality, counter-intuitively, the service provider might speed up the service, which lowers the service quality. In addition, we find that social welfare loss can arise due to the peripheral price or the revenue sharing ratio being strategically determined by an external party (e.g., a peripheral supplier). This loss increases in the peripheral price or the external party's share of the revenue, and it is often most severe when the customers' quality sensitivity is either low or high. Finally, we explore the situation in which a longer service time can reduce peripheral consumption (as can occur in healthcare and repair settings), examining highly debated regulation policies that set price ceilings on the service fee. We find that although such policies can lead to more customers being served, they can also result in both poorer service quality and more peripheral consumption per customer, which reduces social welfare, especially when customers are sensitive to the service quality.

Speed–Quality Trade-Offs in a Dynamic Model

An important trade-off organizations face in many environments is one between quality and speed. Working faster may result in greater output and less delay, but may result in lower quality and dissatisfied customers. In this work, we consider dynamic models in a monopoly setting to explore the optimal balance among the multiple dimensions of speed, price, and wait time. The impact of quality is captured via the market demand potential, which is a function of the speed (quality) in the previous period. We obtain several results and insights. First, in scenarios where speed may be difficult to change over time (e.g., some automated production lines) but price can be changed, we show that the optimal price charged is such that the demand rate remains constant over time, even though the price and market potential are changing. Furthermore, we identify conditions when the firm will work at a speed that is higher or lower than a benchmark speed and characterize the behavior of prices over time. Second, in scenarios where a firm may not be able to change prices but can adjust the speed each period, the firm starts at a speed that may be faster or slower than a benchmark speed but converges to it over time. In this constant price case, as the benchmark speed increases, the initial speed adopted by the firm is actually lower but increases more quickly thereafter. We also characterize the behavior of price and speed in settings where both can be changed over time. Interestingly, a firm typically starts at a slow speed and increases the speed, price, and demand over time. Although our main model assumes that the firm internalizes the congestion cost, several of our results extend to a scenario where the demand rate is impacted by the congestion level.

Balancing credibility, relevance and legitimacy: A critical assessment of trade-offs in science-policy interfaces

To foster strong connections between knowledge and policy action, science–policy interfaces, and the information they produce and exchange, need to be credible, relevant and legitimate. Though this is widely accepted, there has been less emphasis on the problem of trade-offs between these attributes, and how the trade-offs manifest themselves in practice. Based on empirical material on biodiversity-related science–policy interfaces, we identify four major potential trade-offs: first, personal time trade-off: interfacing versus doing other activities; secondly, a clarity–complexity trade-off: simple messages versus communicating uncertainty; thirdly, a speed–quality trade-off: timely outputs versus in-depth quality assessment; and finally, push–pull trade-off: supply-driven versus demand-driven research. Trade-offs are dynamic, vary through policy cycles, and evolve with changing contexts or internal dynamics between actors at the science–policy interface. We outline ways of easing the tensions inherent in trade-offs, but stress that appropriate solutions must be determined on a case-by-case basis.

Analysis of a two echelon Queueing-service system by considering Speed-Quality tradeoff in service

In this paper we study a two-echelon system including a vendor and a service provider in customer-intensive services which introduced by Anand et al. (2011). Customers wish for high quality of service, which leads to longer service time for the service-provider. In the other hand, longer service times can cause to longer waiting time which is not desired for customers. So, tradeoff between service rate and quality of service/or product is critical. The demand is according to Poisson process. The vendor will immediately send the order to the service-provider which works as a M/M/1 production system. The transportation time between the service-provider and vendor controlled by the service-provider, and is assumed to be deterministic and constant. Each customer has a different waiting cost per unit time and different expected value from a product or service; therefore, they decide whether to join the system or not based on the service value, price and expected waiting cost. We derive the speed quality equilibrium of the system as well as the optimal service rate and maximum demand attraction. Furthermore, we investigate the best policy of the system including optimal transportation time, price and service rate. We show that the optimal transportation time is independent from Degree of customer-intensity (9) of the service. Also, the equilibrium price will cause classifying the market, namely market base for small values of 9, and individual base for high values of 9. Unlike in the market base, losing the much share is not undesirable in the individual base.

Effects of electronic performance monitoring on personal control, task satisfaction, and task performance.

Summary of Hypotheses The present study manipulated characteristics of anEPM system that have been the focus of recent legislativeand regulatory attempts. The experiment was designedto test effects of two related issues pertaining to controlin an EPM context. The first issue concerns the worker'sability to prevent or delay monitoring (called monitoringlockout). The second issue concerns whether the workeris made aware of monitoring specifically at each moment(called secrecy). The research measured effects of ma-nipulations of these characteristics on personal control,task satisfaction, and task performance of individualworkers. Personal control in this context is defined asworkers' sense of volitional control over aspects of perfor-mance such as work rate and speed-quality trade-off. Re-cords of workers' performance on a common office com-puter task served as objective measures of the quantityand quality of performance. In addition, workers pro-vided subjective reports of feelings of personal controland task satisfaction. Analyses of these measures wereconducted to test the following hypotheses:A worker's feelings of personal control will be affected bythe design of an electronic monitoring system. When theworker has methods to lock out (delay or prevent) perfor-mance monitoring, feelings of personal control will in-crease (cf. Aiello & Svec, 1993). (Hypothesis la)The worker's feelings of control will also be affected if thesystem informs the worker about the onset of monitoring(i.e., the secrecy factor). In accord with legislative and reg-ulatory activities, we propose that this knowledge of moni-toring will increase feelings of personal control. However,the directionality of this proposition is only equivocallysupported by research findings. Research on warning sig-nals (e.g., Averill, O'Brien, & DeWitt, 1977; Averill & Ro-senn, 1972) indicates that beneficial effects of such signalsdepend both on the nature of the threat and the availabilityof a coping response. (Hypothesis Ib)Task satisfaction will be positively related to feelings of per-sonal control (cf. Greenberger et al., 1989). Higher satis-