Proper Metrics Research Articles

Enhancing Machine Learning-Based Autoscaling for Cloud Resource Orchestration

Performance and cost-effectiveness are sustained by efficient management of resources in cloud computing. Current autoscaling approaches, when trying to balance between the consumption of resources and QoS requirements, usually fall short and end up being inefficient and leading to service disruptions. The existing literature has primarily focuses on static metrics and/or proactive scaling approaches which do not align with dynamically changing tasks, jobs or service calls. The key concept of our approach is the use of statistical analysis to select the most relevant metrics for the specific application being scaled. We demonstrated that different applications require different metrics to accurately estimate the necessary resources, highlighting that what is critical for an application may not be for the other. The proper metrics selection for control mechanism which regulates the requried recources of application are described in this study. Introduced selection mechanism enables us to improve previously designed autoscaler by allowing them to react more quickly to sudden load changes, use fewer resources, and maintain more stable service QoS due to the more accurate machine learning models. We compared our method with previous approaches through a carefully designed series of experiments, and the results showed that this approach brings significant improvements, such as reducing QoS violations by up to 80% and reducing VM usage by 3% to 50%. Testing and measurements were conducted on the Hungarian Research Network (HUN-REN) Cloud, which supports the operation of over 300 scientific projects.

Integrating Surface-related Indicators of Coverage, Distance and Distribution for Quantifying Scan-to-BIM Confidence Level

Abstract. Scan-to-BIM is a widely-used approach to generate Building Information Modelling and by extension Digital Twin models in the architecture, engineering, and construction sector. The resulting models need to be as accurate as possible to ensure subsequent activities that make use of them can do so effectively. Quality assessment of point clouds and occlusion assessment of BIM outputted from Scan-to-BIM has been investigated. However, Scan-to-BIM systems currently do not provide proper metrics as to the confidence the user can have in the quality, in particular geometric quality of the outputted model. This paper addresses this gap by introducing a confidence index, for analysing the reliability of the generated 3D models and thereby quantifying the confidence the user can have in them. Index of confidence is itself derived from three more specific indices: index of coverage estimates the portion of the surface of the modelled element that is explained by the input point cloud. Index of distribution estimates how well the points explaining the modelled surfaces are distributed around the overall object’s surface. Index of distance defines the closeness of the generated element models to the input point cloud. The proposed indices are assessed using three real examples, demonstrating their adequacy.

Lipschitz-free spaces over properly metrizable spaces and approximation properties

Let T T be a topological space admitting a compatible proper metric, that is, a locally compact, separable and metrizable space. Let M T \mathcal {M}^T be the non-empty set of all proper metrics d d on T T compatible with its topology, and equip M T \mathcal {M}^T with the topology of uniform convergence, where the metrics are regarded as functions on T 2 T^2 . We prove that the set A T , 1 \mathcal {A}^{T,1} of metrics d ∈ M T d\in \mathcal {M}^T for which the Lipschitz-free space F ( T , d ) \mathcal {F}(T,d) has the metric approximation property is a dense set in M T \mathcal {M}^T , and is furthermore residual in M T \mathcal {M}^T when T T is zero-dimensional. We also prove that if T T is uncountable then the set A f T \mathcal {A}^T_f of metrics d ∈ M T d\in \mathcal {M}^T for which F ( T , d ) \mathcal {F}(T,d) fails the approximation property is dense in M T \mathcal {M}^T . Combining the last statement with a result of Dalet, we conclude that for any ‘properly metrizable’ space T T , A f T \mathcal {A}^T_f is either empty or dense in M T \mathcal {M}^T .

Two Proper Metrics for Quantifying Remote Synchronization of Identical Oscillator Systems

Remote synchronization (RS) has attracted considerable attention in nonlinear oscillator systems during the past decade. Two metrics, τdirect and τindirect, are proposed in this study to measure RS and identify its associated type in systems of identical oscillators, addressing limitations of traditional indicators used for nonidentical oscillators. We confirm these two metrics' correctness by comparing the identified stable regions with the actual regions of the obtained system through numerical integration, consistent with each other perfectly. Furthermore, these metrics remain valid for systems of nonidentical oscillators and yield conclusions consistent with the traditional Kuramoto order parameters, which are commonly employed to quantify RS. It is expected that the two metrics will enhance the exploration of RS properties in future studies.

Standardized CycleGAN training for unsupervised stain adaptation in invasive carcinoma classification for breast histopathology.

Generalization is one of the main challenges of computational pathology. Slide preparation heterogeneity and the diversity of scanners lead to poor model performance when used on data from medical centers not seen during training. In order to achieve stain invariance in breast invasive carcinoma patch classification, we implement a stain translation strategy using cycleGANs for unsupervised image-to-image translation. Those models often suffer from a lack of proper metrics to monitor and stop the training at a particular point. We also introduce a method to solve this issue. We compare three CycleGAN-based approaches to a baseline classification model obtained without any stain invariance strategy. Two of the proposed approaches use CycleGAN's translations at inference or training to build stain-specific classification models. The last method uses them for stain data augmentation during training. This constrains the classification model to learn stain-invariant features. Regarding CycleGANs' training monitoring, we leverage Fréchet inception distance between generated and real samples and use it as a stopping criterion. We compare CycleGANs' models stopped using this criterion and models stopped at a fixed number of epochs. Baseline metrics are set by training and testing the baseline classification model on a reference stain. We assessed performances using three medical centers with H&E and H&E&S staining. Every approach tested in this study improves baseline metrics without needing labels on target stains. The stain augmentation-based approach produced the best results on every stain. Each method's pros and cons are studied and discussed. Moreover, FID stopping criterion proves superiority to methods using a predefined number of training epoch and has the benefit of not requiring any visual inspection of CycleGAN results. We introduce a method to attain stain invariance for breast invasive carcinoma classification by leveraging CycleGAN's abilities to produce realistic translations between various stains. Moreover, we propose a systematical method for scheduling CycleGANs' trainings by using FID as a stopping criterion and prove its superiority to other methods. Finally, we give an insight on the minimal amount of data required for CycleGAN training in a digital histopathology setting.

Robust Texture-Aware Local Adaptive Image Watermarking With Perceptual Guarantee

Watermarking involves embedding a watermark in an image and later extracting it to prove the image’s copyright. In most cases, a complete image contains both smooth and textured regions. As a rule of thumb, the visual quality of an image with a watermark embedded in its textured regions is better than that of the same image with a watermark in smooth regions. This paper, by taking advantage of the fact, proposes a texture-aware local adaptive watermarking algorithm to maximize the watermark’s robustness while maintaining its imperceptibility. To identify textured regions in an image, we introduce the texture value, an efficient and proper metric of the richness of image texture. It combines the texture correlation of the AC coefficients, the luminance masking of the DC coefficient, and the distribution of image texture. A watermark is embedded adaptively into multiple non-overlapping textured regions of an image under the specified SSIM condition. Its adaptiveness comes from a novel texture-aware adaptive parameter model derived by multivariate regression analysis. Correct extraction of watermarks from multiple textured regions can be done by the cooperation of embedding and extraction strategies, with the assistance of RS-based watermark coding model. They allow for greater robustness, faster extraction, and adjustable watermark capacity. The simulation experiments on 100 images demonstrate that our proposed algorithm outperforms state-of-the-art algorithms with respect to imperceptibility, robustness, and adaptability.

Measuring sustainable urban development in residential areas of the 20 biggest Finnish cities

As a result of the ongoing urbanization megatrend, cities have an increasingly critical role in the search for sustainability. To create sustainable strategies for cities and to follow up if they induce desired effects proper metrics on the inter and intra-urban development is needed. In this paper, we analyze the sustainability development in the 20 largest cities in Finland through a residential area classification framework. The results based on high-quality register data show concerning trends in some sustainability measures, and divergent trends between cities and residential areas within. Overall, while densities have increased modestly, we see no clear signs of decreasing car ownership rates. Further, also manifestations of social sustainability seem to be insufficient in many locations–especially in residential mid-rise areas from the '60s and '70s, and '80s and '90s.

Classical models may be a better explanation of the Jiuzhang 1.0 Gaussian Boson Sampler than its targeted squeezed light model

Recently, Zhong et al. [1, 2] performed landmark Gaussian boson sampling experiments with up to 144 modes using threshold detectors. The authors claim to have achieved quantum computational advantage with the implementation of these experiments, named Jiuzhang 1.0 and Jiuzhang 2.0. Their experimental results are validated against several classical hypotheses and adversaries using tests such as the comparison of statistical correlations between modes, Bayesian hypothesis testing and the Heavy Output Generation (HOG) test. In this work we propose an alternative classical hypothesis for the validation of these experiments. We use the probability distribution of mixtures of coherent states sent into a lossy interferometer; these input mixed states, which we term squashed states, have vacuum fluctuations in one quadrature and excess fluctuations in the other. We find that for configurations in the high photon number density regime, the comparison of statistical correlations does not tell apart the ground truth of the experiment (two-mode squeezed states sent into an interferometer) from our alternative hypothesis. On the other hand, the Bayesian test indicates that, for all configurations excepting Jiuzhang 1.0, the ground truth is a more likely explanation of the experimental data than our alternative hypothesis. A similar result is obtained for the HOG test: for all configurations of Jiuzhang 2.0, the test indicates that the experimental samples have higher ground truth probability than the samples obtained from our alternative distribution; for Jiuzhang 1.0 the test is inconclusive. Our results provide a new hypothesis that should be considered in the validation of future GBS experiments, and shed light into the need to identify proper metrics to verify quantum advantage in the context of threshold GBS. Additionally, they indicate that a classical explanation of the Jiuzhang 1.0 experiment, lacking any quantum features, has not been ruled out.

Multi-criteria comparison tools to evaluate cost- and eco-efficiency of ultra-high-performance concrete

This work was motivated by the increasing need for proper metrics and tools to demonstrate the effect of mechanical performance, as a function of concrete mix composition, in dictating the dimensions of structural elements and associated costs and embodied carbon dioxide (CO2) emissions. Mixture compositions associated with different concrete technologies were compared using multi-criteria comparison indices derived using structural design considerations and calculated using information on compressive strength, volumetric embodied CO2 and unit costs. In addition, predicted compressive strengths obtained with machine learning (ML) models are used to calculate these indices for a domain of mix proportions associated with ultra-high-performance concrete materials to generate multi-objective density diagrams (MODDs). The makeup of this tool facilitates the evaluation of rather complicated trends associated with mix proportions and multi-objective outcomes, allowing ML-based tools to be of easy interpretation by industry personnel with no expertise in artificial intelligence. MODDs could be used as aids in the decision-making process during mix design stages and serve as proof of mixture optimization that could be introduced in environmental product declarations. Results show that, in contrast to conventional wisdom, high-binder content and ultra-high strength concrete technologies are not necessarily detrimental to cost and/or eco efficiencies. For the applications evaluated herein, optimum solutions were mostly obtained with these types of concrete, suggesting that industry trends toward requiring minimization of embodied carbon footprint on a per volume of concrete basis are misguided and should not be used as a standalone metric to minimize the total carbon footprint of concrete structures.

Onto Quantifying Unsteady Propulsion Characteristics Using Momentum and Energy Control Volume Assessments

AbstractThis effort presents a novel approach to interrogate efficiency for unsteady, undulating propulsion using variable momentum and energy conservation (VMEC) assessments. These integral approaches utilize large amounts of data from computational fluid dynamics (CFD) to address present difficulties associated with separating thrust from drag associated with propelling bodies as well as potentially resolve issues associated with defining a nonzero efficiency for a body in self-propulsion. Such a fundamental issue is addressed through strategic control volume assessments of the momentum and energy conservation equations. In this work, the Method of Manufactured Solutions (MMS) is used to verify the integral-based evaluation approach and better quantify output. The MMS results indicate the method is valid and that one can separate work associated with lift and drag from the energy budget. This separation procedure provides a means to separate propulsive and drag forces. The effort then studies previously validated CFD simulations of heaving and pitching foils to provide insight associated with separating axial forces into their thrust and drag components for highly complex systems. The effort then presents a new efficiency metric that can obtain nonzero efficiencies in self-propulsion. Overall, the results indicate that energy-based assessments provide insight that is a step forward toward isolating loss from propulsive mechanisms and developing proper metrics of efficiency.

Motion Smoothness-Based Assessment of Surgical Expertise: The Importance of Selecting Proper Metrics.

The smooth movement of hand/surgical instruments is considered an indicator of skilled, coordinated surgical performance. Jerky surgical instrument movements or hand tremors can cause unwanted damages to the surgical site. Different methods have been used in previous studies for assessing motion smoothness, causing conflicting results regarding the comparison among surgical skill levels. We recruited four attending surgeons, five surgical residents, and nine novices. The participants conducted three simulated laparoscopic tasks, including peg transfer, bimanual peg transfer, and rubber band translocation. Tooltip motion smoothness was computed using the mean tooltip motion jerk, logarithmic dimensionless tooltip motion jerk, and 95% tooltip motion frequency (originally proposed in this study) to evaluate their capability of surgical skill level differentiation. The results revealed that logarithmic dimensionless motion jerk and 95% motion frequency were capable of distinguishing skill levels, indicated by smoother tooltip movements observed in high compared to low skill levels. Contrarily, mean motion jerk was not able to distinguish the skill levels. Additionally, 95% motion frequency was less affected by the measurement noise since it did not require the calculation of motion jerk, and 95% motion frequency and logarithmic dimensionless motion jerk yielded a better motion smoothness assessment outcome in distinguishing skill levels than mean motion jerk.

A mathematical characterization of anatomically consistent blood capillary networks

Blood microcirculation is the site of control of tissue perfusion, blood-tissue exchange, and tissue blood volume. Despite the many irregularities, almost ubiquitously, one can recognize in microcirculation vessels a hierarchy of arterioles and venules, organized in tree-like structures, and capillary plexi, organized in net-like structures. Whilst for arterioles and venules it may be envisageable to obtain geometries needed for numerical simulations from imaging techniques, the size and numerosity of capillaries makes this task much more cumbersome. For this reason, it is interesting to study approaches to generate in silico-derived artifacts of capillary networks, even in view of machine-learning based approaches which require a large amount of samples for training. Artificial networks must correctly reflect proper metrics and topology, which in turn, will ensure with proper boundary conditions a physiological blood flux in the net and a sufficient nutrient distribution in the surrounding tissues. In this paper, we introduce the sequence of curves whose limit is the space filling Hilbert curve and we discuss its inherent properties and we obtain the backbone of the artificial capillary network from a suitable element of this sequence. The backbone represents a significant synthesis of basic metric features of the network and, in this context, its properties can be studied analytically. In this framework, the Hilbert curve is a malleable entity which allows to shape the backbone according to the physical indicators. In particular, two significant factors are shown to control the network topology and scaling: the iteration step of the Hilbert curve generation and the characteristic length of the REV, respectively. Based on the points we generate for a certain iteration step, we then obtain via spline interpolation a smoothed version of the curve, which fine–tunes the tortuosity. A volumetric construction is obtained building a tubular neighborhood of the backbone, whose metrics can be computed and tuned as well. Numerical simulations of the blood flow in the obtained geometry show the physical fields occurring in the artificial network.

On measuring coupling between microservices

In software quality management, the selection strategy for proper metrics varies depending on the application scenarios and measurement objectives. MicroService Architecture (MSA), despite being commonly employed nowadays, still cannot be reliably measured and compared if the microservices in a system are independent. Software managers and architects need to understand whether their microservices are “decoupled enough”, if not, which ones are over-coupled, and by how much. In this paper, we contribute a novel set of metrics – Microservice Coupling Index (MCI) – derived from the relative measurement theory. Instead of measuring coupling evidence with simple counts, we measure how dependent and coupled the microservices are relative to the possible couplings between them. We measured the MCI metrics for 15 open source projects that involve 113 distinct microservices. Empirical investigation confirmed that MCIs differ quite significantly from existing coupling measures and that they are more discriminative than existing ones for separating high and low degrees of microservice couplings and thus more useful in comparing design alternatives. A series of experimental studies were conducted, showing that the larger the MCIs, the less likely the bugs and changes can be localized and separated, and the less likely that the individual microservices in a system can be independently developed and evolved.

Discriminating abilities of threshold-free evaluation metrics in link prediction

Link prediction is a paradigmatic and challenging problem in network science, which attempts to uncover missing links or predict future links, based on known topology. A fundamental but still unsolved issue is how to choose proper metrics to fairly evaluate prediction algorithms. The area under the receiver operating characteristic curve (AUC) and the balanced precision (BP) are the two most popular metrics in early studies, while their effectiveness is recently under debate. At the same time, the area under the precision–recall curve (AUPR) becomes increasingly popular, especially in biological studies. Based on a toy model with tunable noise and predictability, we propose a method to measure the discriminating ability of any given metric. We apply this method to the above three threshold-free metrics, showing that AUC and AUPR are remarkably more discriminating than BP, and AUC is slightly more discriminating than AUPR. The result suggests that it is better to simultaneously use AUC and AUPR in evaluating link prediction algorithms. At the same time, it warns us that the evaluation based only on BP may be unauthentic. This article provides a starting point towards a comprehensive picture about effectiveness of evaluation metrics for link prediction.

Distinctive Measurement Scheme for Security and Privacy in Internet of Things Applications Using Machine Learning Algorithms

More significant data are available thanks to the present Internet of Things (IoT) application trend, which can be accessed in the future using some platforms for data storage. An external storage space is required for practical purposes whenever a data storage platform is created. However, in the IoT, certain cutting-edge storage methods have been developed that compromise the security and privacy of data transfer processes. As a result, the suggested solution creates a standard mode of security operations for storing the data with little noise. One of the most distinctive findings in the suggested methodology is the incorporation of machine learning algorithms in the formulation of analytical representations. The aforementioned integration method ensures high-level quantitative measurements of data security and privacy. Due to the transmission of large amounts of data, users are now able to assess the reliability of data transfer channels and the duration of queuing times, where each user can separate the specific data that has to be transferred. The created system is put to the test in real time using the proper metrics, and it is found that machine learning techniques improve security more effectively. Additionally, for 98 percent of the scenarios defined, the accuracy for data security and privacy is maximized, and the predicted model outperforms the current method in all of them.

Extending proper metrics

We first prove a version of Tietze-Urysohn's theorem for proper functions taking values in non-negative real numbers defined on σ-compact locally compact Hausdorff spaces. As an application, we prove an extension theorem of proper metrics, which states that if X is a σ-compact locally compact Hausdorff space, A is a closed subset of X, and d is a proper metric on A that generates the same topology of A, then there exists a proper metric D on X such that D generates the same topology of X and D|A2=d. Moreover, if A is a proper retract, we can choose D so that (A,d) is quasi-isometric to (X,D). We also show analogues of the theorems explained above for ultrametrizable spaces.

The effect of non-uniform urban illumination upon pedestrians’ alertness and sense of safety

The influence of urban lighting on the visual tasks of pedestrians are well known and, likewise, studies have explored the relationship between urban lighting and pedestrians’ feelings of safety. But there are other, less investigated, possible effects of light in mesopic situations. Research on street lighting for pedestrians has traditionally focused on minimal required illuminance for various visual performance tasks (e.g., obstacle detection and face recognition). However, it is not only pedestrians’ visual performance that is important for safety, but also their general attentiveness to the environment. Despite this, the psychological concepts related to the individuals’ attention, such as alertness, arousal and anxiety, have not been considered in street lighting research to date even though they too can play an important role in pedestrian’s attention and safety. Since proper metrics for these concepts are needed, we conducted a pilot study to test their sensitivity. We decided for a conceptual replication of a field experiment from 1916 and managed to partly replicate the findings. More importantly, we learned meaningful lessons from this endeavour. Implications for future research are discussed.

Distance distributions and inverse problems for metric measure spaces

AbstractApplications in data science, shape analysis, and object classification frequently require comparison of probability distributions defined on different ambient spaces. To accomplish this, one requires a notion of distance on a given class of metric measure spaces—that is, compact metric spaces endowed with probability measures. Such distances are typically defined as comparisons between metric measure space invariants, such as distance distributions (also referred to as shape distributions, distance histograms, or shape contexts in the literature). Generally, distances defined in terms of distance distributions are actually pseudometrics, in that they may vanish when comparing nonisomorphic spaces. The goal of this paper is to set up a formal framework for assessing the discrimininative power of distance distributions, that is, the extent to which these pseudometrics fail to define proper metrics. We formulate several precise inverse problems in terms of these invariants and answer them in several categories of metric measure spaces, including the category of plane curves, where we give a counterexample to the curve histogram conjecture of Brinkman and Olver, the categories of embedded and Riemannian manifolds, where we obtain sphere rigidity results, and the category of metric graphs, where we obtain a local injectivity result along the lines of classical work of Boutin and Kemper on point cloud configurations. The inverse problems are further contextualized by the introduction of a variant of the Gromov–Wasserstein distance on the space of metric measure spaces, which is inspired by the original Monge formulation of optimal transport.

Pursuing occupants' health and well-being in building management: Definition of new metrics based on indoor air parameters.

The spread of COVID-19 has affected the lives of millions of people. Pandemic has made people more sensitive to health issues. In particular, the growing concern for the virus spread in confined spaces has promoted the necessity to improve indoor air quality. Literature is stressing how buildings must be designed and operated pursuing occupants’ health and well-being, with a particular attention for indoor air parameters. This poses the challenge of monitoring and assessing these aspects through proper metrics. In this paper the approach towards a multi-step assessment procedure embedding in buildings assessment health and well-being related variables and indicators is elaborated. They are intended to inform a building manager of the potential influence of air conditions on human health and well-being. Moreover, a set of monetary metrics (i.e., impacts) is proposed to translate energy and indoor air related building performances into euros, putting the basis for a comprehensive economic evaluation. The application of the set of proposed metrics to an Italian hotel (i.e., Italian pilot of H2020 MOBISTYLE project), enabled to map some indoor air conditions causing health concerns, and to identify clusters of guests with best and worst indoor air conditions, to be targeted by new management strategies. Despite case study specific limitations, the application exemplified how the methodology can expand the traditional energy-based performance assessment for building management towards indoor air domain and the related economic impacts, with implication on results in terms of overall economic performance of the building from both a private and public perspective.

A unified framework on defining depth for point process using function smoothing

The notion of statistical depth has been extensively studied in multivariate and functional data over the past few decades. In contrast, the depth on temporal point process is still under-explored. The problem is challenging because a point process has two types of randomness: 1) the number of events in a process, and 2) the distribution of these events. Recent studies proposed depths in a weighted product of two terms, describing the above two types of randomness, respectively. Under a new framework through a smoothing procedure, these two randomnesses can be unified. Basically, the point process observations are transformed into functions using conventional kernel smoothing methods, and then the well-known functional h-depth and its modified, center-based version are adopted to describe the center-outward rank in the original data. To do so, a proper metric is defined on the point processes with smoothed functions. Then an efficient algorithm is provided to estimate the defined “center”. The mathematical properties of the newly defined depths are explored and the asymptotic theories are studied. Simulation results show that the proposed depths can properly rank point process observations. Finally, the new methods are demonstrated in a classification task using a real neuronal spike train dataset.