Data Point Research Articles

Black-box tests for algorithmic stability.

Algorithmic stability is a concept from learning theory that expresses the degree to which changes to the input data (e.g. removal of a single data point) may affect the outputs of a regression algorithm. Knowing an algorithm's stability properties is often useful for many downstream applications-for example, stability is known to lead to desirable generalization properties and predictive inference guarantees. However, many modern algorithms currently used in practice are too complex for a theoretical analysis of their stability properties, and thus we can only attempt to establish these properties through an empirical exploration of the algorithm's behaviour on various datasets. In this work, we lay out a formal statistical framework for this kind of black-box testing without any assumptions on the algorithm or the data distribution, and establish fundamental bounds on the ability of any black-box test to identify algorithmic stability.

Exploring the clinical relevance of vital signs statistical calculations from a new-generation clinical information system

New information on the intensive care applications of new generation ‘high-density data clinical information systems’ (HDDCIS) is increasingly being published in the academic literature. HDDCIS avoid data loss from bedside equipment and some provide vital signs statistical calculations to promote quick and easy evaluation of patient information. Our objective was to study whether manual records of continuously monitored vital signs in the Paediatric Intensive Care Unit could be replaced by these statistical calculations. Here we conducted a prospective observational clinical study in paediatric patients with severe diabetic ketoacidosis, using a Medlinecare® HDDCIS, which collects information from bedside equipment (1 data point per parameter, every 3–5 s) and automatically provides hourly statistical calculations of the central trend and sample dispersion. These calculations were compared with manual hourly nursing records for patient heart and respiratory rates and oxygen saturation. The central tendency calculations showed identical or remarkably similar values and strong correlations with manual nursing records. The sample dispersion calculations differed from the manual references and showed weaker correlations. We concluded that vital signs calculations of central tendency can replace manual records, thereby reducing the bureaucratic burden of staff. The significant sample dispersion calculations variability revealed that automatic random measurements must be supervised by healthcare personnel, making them inefficient.

An Application of K-Nearest-Neighbor Regression in Maize Yield Prediction

Predictive analytics utilizes historical data and knowledge to predict future outcomes and provides a method for evaluating the accuracy and reliability of these forecasts. Artificial Intelligence is a tool of predictive analytics. AI trains computers to learn human behaviors like learning, judgment, and decision-making while simulating intelligent human behavior using computers and has received a lot of attention in almost all areas of research. Machine learning is a branch of Artificial Intelligence that has been used to solve classification and regression problems. Machine learning advancements have aided in boosting agricultural gains. Yield prediction is one of the agricultural sectors that has embraced machine learning. K Nearest Neighbor (KNN) Regression is a regression algorithm used in machine learning for prediction tasks. KNN Regression is like KNN Classification, except that KNN Regression predicts a constant output value for a given input instead of predicting a class label. The basic idea behind KNN Regression is to find the K nearest neighbors to a given input data point based on a distance metric and then use the average (or weighted average) of the output values of these K neighbors as the predicted output for the input data point. The distance metric used in KNN Regression can vary depending on the data type being analyzed, but common distance metrics include Euclidean distance, Manhattan distance, and Minkowski distance. This paper presents the application of KNN regression in maize yield prediction in Uasin Gishu county, in north rift region of Kenya. Questionnaires were distributed to 900 randomly selected maize farmers across the thirty wards to obtain primary data. With a Train Test split ration of 80:20, KNN regression algorithm was able to predict maize yield and its prediction performance was evaluated using Root Mean Squared error-RMSE=0.4948, Mean Squared Error-MSE =0.2803, Mean Absolute Error-MAE = 0.4591 and Mean Absolute Percentage Error-MAPE = 36.17. According to the study findings, the algorithm was able to predict maize yield in the maize producing county.

Reproductive, maternal, newborn, and child health intervention coverage in 70 low-income and middle-income countries, 2000–30: trends, projections, and inequities

Monitoring the progress in reproductive, maternal, newborn, and child health (RMNCH) using the composite coverage index (CCI) is crucial to evaluate the advancement of low-income and middle-income countries (LMICs) towards the attainment of Sustainable Development Goal target 3. We present current benchmarking for 70 LMICs, forecasting to 2030, and an analysis of inequities within and across countries. In this cross-sectional secondary data analysis, we extracted 291 data points from the WHO Equity Monitor, and Demographic and Health Survey Statcompiler for 70 LMICs. We selected countries on the basis of whether they belonged to LMICs, had complete information about the predictors between 2000 and 2030, and had at least one data point related to CCI. CCI was calculated on the basis of eight types of RMNCH interventions in four domains, comprising family planning, antenatal care, immunisations, and management of childhood illnesses. This study examined CCI as the main outcome variable. Bayesian hierarchical models were used to estimate trends and projections of the CCI at regional and national levels, as well as the area of residence, educational level, and wealth quintile. Despite progress, only 18 countries are projected to reach the 80% CCI target by 2030. Regionally, CCI is projected to increase in all regions of Asia (in southern Asia from 51·8% in 2000 to 89·2% in 2030; in southeastern Asia from 58·8% to 84·4%; in central Asia from 70·3% to 87·0%; in eastern Asia from 76·8% to 82·1%; and in western Asia from 56·5% to 72·1%), Africa (in sub-Saharan Africa from 46·3% in 2000 to 72·2% in 2030 and in northern Africa from 55·0% to 81·7%), and Latin America and the Caribbean (from 67·0% in 2000 to 83·4% in 2030). By contrast, southern Europe is predicted to experience a decline in CCI over the same period (70·1% in 2000 to 55·2% in 2030). Across LMICs, CCIs are higher in urban areas, in populations in which women have higher education levels, and in populations with a high income. Governments of countries where the universal target of 80% CCI has not yet been reached must develop evidence-based policies aimed at enhancing RMNCH coverage. Additionally, they should focus on reducing the extent of existing inequalities within their populations to drive progress in RMNCH. Hitotsubashi University and Japan Society for the Promotion of Science.

Hesitant fuzzy time series forecasting model of higher order based on one and two-factor aggregate logical relationship

Fuzzy time series forecasting is a methodology that incorporates fuzzy logic to handle uncertainty in time series data. It allows for more flexible and accurate predictions by considering the linguistic terms and membership functions associated with each data point, enabling a better representation of complex patterns and trends within the time series. A time series forecasting model examines the relationships between present and past observations for future measurement. This article proposes a hesitant fuzzy time series forecasting (FTSF) technique for higher order based on one and two-factor aggregate fuzzy relationships. The proposed method presents a new approach to partitioning intervals into an unequal and equal length. The hesitant fuzzy sets are defined for each observation, and fuzzification is based on their maximum score function. Further, aggregated triangular fuzzy numbers are defined using an aggregation operator corresponding to the fuzzified datum. In order to enhance forecasting performance, this work also introduces a novel defuzzification rule based on aggregated fuzzy logical relationship groups (AFLRGs) for higher order one and two-factor fuzzy time series. The implementation of the proposed approach is verified on two data sets to the forecasting of enrollment data and TAIFEX and TAIEX stock index data sets, and the forecasting error is tested in term of different statistical parameters. Thus, the model validates the effectiveness of the proposed method with higher forecasting accuracy rate in the environment hesitant fuzzy information.

An Entropy Weighted Nonnegative Matrix Factorization Algorithm for Feature Representation.

Nonnegative matrix factorization (NMF) has been widely used to learn low-dimensional representations of data. However, NMF pays the same attention to all attributes of a data point, which inevitably leads to inaccurate representations. For example, in a human-face dataset, if an image contains a hat on a head, the hat should be removed or the importance of its corresponding attributes should be decreased during matrix factorization. This article proposes a new type of NMF called entropy weighted NMF (EWNMF), which uses an optimizable weight for each attribute of each data point to emphasize their importance. This process is achieved by adding an entropy regularizer to the cost function and then using the Lagrange multiplier method to solve the problem. Experimental results with several datasets demonstrate the feasibility and effectiveness of the proposed method. The code developed in this study is available at https://github.com/Poisson-EM/Entropy-weighted-NMF.

Generative diffusion in very large dimensions

Generative models based on diffusion have become the state of the art in the last few years, notably for image generation. Here, we analyze them in the high-dimensional limit, where data are formed by a very large number of variables. We use methods from statistical physics and focus on two well-controlled high-dimensional cases: a Gaussian model and the Curie–Weiss model of ferromagnetism. In the latter case, we highlight the mechanism of symmetry breaking in the inverse diffusion, and point out that, in order to reconstruct the relative asymmetry of the two low-temperature states, and thus to obtain the correct probability weights, one needs a database with a number of points much larger than the dimension of each data point. We characterize the scaling laws in the number of data and in the number of dimensions for an efficient generation.

410. IMPACT ON BODY WEIGHT OF REMNANT STOMACH PRESERVING ESOPHAGECTOMY FOR THE PATIENTS WITH PAST MEDICAL HISTORY OF GASTRECTOMY

Abstract Background Esophagectomy has a great impact on patients’ nutritional status. Particularly for the patients with past medical history of gastrectomy, there is a concern that surgical procedure of esophagectomy accommpanying remnant gastrectomy may potentially deteriorate nutritional status more than standard esophagectomy, resulting in excessive body weight loss. In order to reduce such negative impact, we perform remnant stomach preserving esophagectomy. The purpose of this study is to assess the impact on body weight of this procedure. Methods The patients with past medical history of gastrectomy who underwent remnant stomach preserving esophagectomy in our institute from 2015 to 2022 were enrolled in this study. All the measured body weights were collected from the electronic medical record retrospectively and all data point were plotted as BMI over time to see any transitional trend. The day of esophagectomy was defined as day0. The enrolled patients were divided into two groups according to the median initial BMI; lower BMI group and higher BMI group. The BMI transitional trends were detected in each group. Results Totally 37 patients were eligible in this study. Median age was 72 (57–88), the number of female was 1 (36 male), cStage1/2/3/4 of esophageal cancer were 9/9/13/5, median time period from gastrectomy was 21.0 years (0–54.5), and median initial BMI was 20.7 (15.1–27.7). All BMI data points from every patient according to the higher/lower initial BMI groups were shown in the figure. The decreasing trend in BMI until about 3 months after the esophagectomy was seen in the higher BMI group whilst relatively plateau trend in such period was seen in the lower BMI group. Conclusion Efficacy of the remnant stomach preserving esophagectomy in terms of reducing body weight loss could be more apparent in lower BMI patients than higher BMI patients. When you see an esophageal cancer patient with past medical history of gastrectomy, the remnant stomach preserving procedure should be considered particularly when the BMI is low.

Defining Clusters by Topology Warping Features, an Interpretable Data Clustering Method

Clustering is the task of dividing a data-set into different groups, called clusters, based on similarity. Despite being extensively studied, many state-of-the-art clustering algorithms lack interpretability. While it is useful to partition objects into clusters, it can be equally useful to understand why each cluster has been created. This motivates the desire for a clustering algorithm which can explain its partition. Our proposed method is a clustering process which seeks to explain which variables of a data-set are responsible for each cluster. We analyze the shape of the data, through the mathematical concept of a topological space. A space which is optimal for clustering is one which contains several disconnected islands, quantified as the number of connected components. Subsets of variables can then be selected and the resulting connected components of the topological space calculated. If this space is promising then the resulting disconnected region is the set of data-points which make up a cluster, and the subset of variables are what define it. We chose the variables to consider by grouping them via their correlation or through complex methods, e.g evolutionary algorithms. Our method provides a simple explanation since we can confidently assert that particular variables are why a data point is in a certain cluster. Alongside test data for comparison with existing algorithms, our methodology was applied to a community-based adolescents lipidomics dataset. Results on this dataset revealed 3 distinct clusters which can be explained by the distinct set of lipids that define them. Further optimization showed the existence of a cluster defined by a single lipid.

Application of Fuzzy C-Means Algorithm for Clustering Customers

Today's fierce business competition requires companies to focus on the needs desired by consumers.This makes companies have to think about how to manage customer data so that it can be utilized properly for the development of marketing strategies.And Grouping (cluster) customers based on their respective characteristics can be an alternative in solving these problems.In clustering (cluster) customers there are several data mining clustering methods that can be used, one of which is the Fuzzy C-Means (FCM) method. FCM is a clustering algorithm where one object can be a member of several clusters and FCM cluster boundaries are vague. The output of FCM is a row of cluster centers and several degrees of membership for each data point. In this clustering, customers will be divided into 4 customer clusters namely Golden, Silver, Bronze, and Iron with the variables used as a reference are the final purchase date, purchase frequency and total purchase. The data used is customer transaction data for the period September - December 2015. The total data is 709 transactions from 75 customers. After the data is processed with the Fuzzy C-Means method, the final results show that the iteration ends at the 30th iteration with a change in the objective function of 9.8. The resulting customer clusters are Golden: 27, Silver: 15, and Bronze: 33 with a cluster validity of 0.596277.

A Comparison of Decision Rule Accuracy From Curriculum-Based Measurement of Reading and Nonsense Word Fluency

The transition from sounding out unfamiliar words to effortlessly reading connected text does not occur all at once nor at the same rate for students. The purpose of this study was to explore the accuracy of three decision rules (data point, median, and trend line) applied to progress monitoring outcomes of alphabetic principle (nonsense word fluency [NWF]) and oral reading rate (curriculum-based measurement of reading [CBM-R]). Outcomes from students receiving Tier-2 supports in oral reading and decoding were analyzed to generate model parameters. Scores were simulated for NWF and CBM-R and decision rules were applied to schedules where one observation was collected per week. The trend-line rule was viable with NWF after 7 weeks and 9 to 10 weeks with CBM-R. Differences in base rates of non-proficiency between measures call into question the utility of NWF to capture student improvement in alphabetic principle as they encounter increasingly complex word types.

Quantum State Assignment Flows.

This paper introduces assignment flows for density matrices as state spaces for representation and analysis of data associated with vertices of an underlying weighted graph. Determining an assignment flow by geometric integration of the defining dynamical system causes an interaction of the non-commuting states across the graph, and the assignment of a pure (rank-one) state to each vertex after convergence. Adopting the Riemannian-Bogoliubov-Kubo-Mori metric from information geometry leads to closed-form local expressions that can be computed efficiently and implemented in a fine-grained parallel manner. Restriction to the submanifold of commuting density matrices recovers the assignment flows for categorical probability distributions, which merely assign labels from a finite set to each data point. As shown for these flows in our prior work, the novel class of quantum state assignment flows can also be characterized as Riemannian gradient flows with respect to a non-local, non-convex potential after proper reparameterization and under mild conditions on the underlying weight function. This weight function generates the parameters of the layers of a neural network corresponding to and generated by each step of the geometric integration scheme. Numerical results indicate and illustrate the potential of the novel approach for data representation and analysis, including the representation of correlations of data across the graph by entanglement and tensorization.

Robust embedding regression for semi-supervised learning

To utilize both labeled data and unlabeled data in real-world applications, semi-supervised learning is widely used as an effective technique. However, most semi-supervised methods do not perform well when there are many noises and redundant information in the original data. To address these issues, in this paper, we proposed a novel approach called robust embedding regression (RER) for semi-supervised learning by inheriting the advantages of the existing semi-supervised learning, robust linear regression, and low-rank representation techniques. Specifically, RER constructs a more robust and accurate graph by adaptively arranging the weight coefficient for each data point. Furthermore, the low-rank representation is introduced to reduce the negative influence of the redundant features and noises residing in the original data while the graph construction. More importantly, the proper norms are imposed on both the reconstruction and regularization terms to further improve the robustness and earn feature/sample selection. We designed an effective iterative algorithm to optimize the problem of RER. Comprehensive experimental results conducted on both synthetic and real-world datasets indicate that RER is superior in classification and clustering performance and robust to different types of noise compared with the existing semi-supervised methods.

Attribute Selection Algorithm with Clustering based Optimization Approach based on Mean and Similarity Distance

With hundreds or thousands of attributes in high-dimensional data, the computational workload is challenging. Attributes that have no meaningful influence on class predictions throughout the classification process increase the computing load. This article's goal is to use attribute selection to reduce the size of high-dimensional data, which will lessen the computational load. Considering selected attribute subsets that cover all attributes. As a result, there are two stages to the process: filtering out superfluous information and settling on a single attribute to stand in for a group of similar but otherwise meaningless characteristics. Numerous studies on attribute selection, including backward and forward selection, have been undertaken. This experiment and the accuracy of the categorization result recommend a k-means based PSO clustering-based attribute selection. It is likely that related attributes are present in the same cluster while irrelevant attributes are not identified in any clusters. Datasets for Credit Approval, Ionosphere, Annealing, Madelon, Isolet, and Multiple Attributes are employed alongside two other high-dimensional datasets. Both databases include the class label for each data point. Our test demonstrates that attribute selection using k-means clustering may be done to offer a subset of characteristics and that doing so produces classification outcomes that are more accurate than 80%.

Neoarchean synkinematic metamorphic peak in the Isua supracrustal belt (West Greenland)

Abstract We present petrological data and seven Lu-Hf garnet–amphibole–whole rock ages obtained from a single garnet-hornblende-mica schist sample from the Isua supracrustal belt (West Greenland). Garnets grew during prograde metamorphism toward regional amphibolite-facies peak conditions, and a mylonitic foliation formed during and after garnet growth. Garnet crystals show typical prograde zoning with no visible traces of a relict garnet generation. They do show various degrees of retrogression. While some crystals are perfectly euhedral with only minor chemical alteration along cracks, others are elongated in the foliation and either grew in this shape or were deformed. Six garnet splits were separated from crushed single crystals and one from a crushed bulk sample. Individual three-point garnet–hornblende–whole rock ages scatter between 2.603 ± 0.018 Ga and 2.432 ± 0.059 Ga for single garnets. The garnet split from the bulk sample defines an age of 2.463 ± 0.031 Ga, the data point farthest from the regression line for all data points (2.551 ± 0.074 Ga, mean square of weighted deviates = 25). We interpret these data to indicate partial retrogression of a Neoarchean garnet population not significantly older than the oldest obtained three-point age. Well-preserved garnet zoning, regional peak temperatures well below the closing temperature of the Lu-Hf system, and the small scatter of Lu-Hf ages preclude an interpretation of the observed metamorphism and deformation as being Eoarchean in age.

Separation of sulfur compounds from petroleum using hydrodesulfurization method to reduce SO2 emission: Process optimization and validation

Preparation of clean fuels from petroleum need decent removal of sulfur compounds from the petroleum which can be done using the well-known hydrodesulfurization (HDS) method. The removal of sulfur compounds from crude oil would be essential in order to control the emission of gas pollutants such as SO2. However, due to the complexity of the sulfur separation process, predictive modeling and computations are required to improve the separation efficiency. In this research, we have collected some experimental data for optimization of separation process. Each data point has four input features: temperature, pressure, initial sulfur content, and dose. On the other hand, the outputs included left sulfur amount in the feed, value of SO2 emission, and process cost ($). Adaboost technique integrated with three core models of DT, Lasso, and KNN is used for modeling. The models are tuned using LOA method on the available dataset, thereby the optimum models’ parameters were determined for the best fitting. For sulfur concentration and emission parameters, the ADA-DT technique is the best one among other methods, but for the HDS cost, the ADA-LASSO framework works the best. Using these models, the R2-score for outputs is, respectively, 0.940, 0.923, and 0.999. This work makes significant contributions by providing accurate predictive models that would enhance the understanding and optimization of sulfur separation process. These models pave the way for more efficient and environmentally-friendly production of clean fuels from petroleum, contributing to reduced gas pollutants emissions.

Predicting ionic liquid efficiency as steel corrosion inhibitor under static hydrochloric acid immersion using group contribution method

As a novel corrosion inhibitor, ionic liquid has garnered a growing amount of interest. Different electrodes, such as carbon steel, stainless steel, aluminium, copper, magnesium, and others; different electrolytes, such as hydrochloric acid, sulfuric acid, nitric acid, and seawater; different condition operations at different temperatures and tribological properties; and various techniques, such as weight loss (WL), potentiodynamic polarization (PDP), and electro impedance spectroscopy (EIS), have been investigated. Due to the large number of cations and anions, this conventional laboratory study may not be completed anytime soon. Therefore, IL as a corrosion inhibitor must be investigated further. The purpose of this study is to conduct a more comprehensive analysis of IL as a corrosion inhibitor in 1 M HCl immersion on carbon steel using the group contribution method. Cations and anions were divided into 29 functional groups (FG) of cations and 17 functional groups (FG) of anions using 74 IL and 760 WL data points collected from various studies. Training and test set results have R2 values of 0.639 and AARD values of 14.9%; validation set results have R2 values of 0.319 and AARD values of 14.4%; and the overall data point evaluation has R2 values of 0.324 and AARD values of 14.3%.

Classification model-based and assisted environment selection for evolutionary algorithms to solve high-dimensional expensive problems

Surrogate-assisted evolutionary algorithms (SAEAs) have been proven to be very effective in tackling low-dimensional expensive problems. However, it remains a challenge to solve high-dimensional expensive problems (HEPs) with the curse of dimensionality. Therefore, this paper proposes a surrogate-assisted improved multioperator differential evolution (SA-IMODE) algorithm to address HEPs up to 2000 dimensions. Specifically, this paper proposes a novel relationship classification model-based environment selection strategy (RCES), in which a classification model is used to distinguish between “good” and “bad” solutions to assist in environment selection. By doing this, the “unpromising” solutions are thrown away directly without evaluation to reduce the number of expensive fitness evaluations. Moreover, the solution-improvement data point and error samples are added to construct the training dataset to improve the model's prediction accuracy. Two coordinate systems are utilized to produce five types of DE operators to suit different fitness landscapes. An adaptive strategy is also used to select suitable DE operators to generate promising offspring. Furthermore, a local search mechanism is applied to refine the best solution in the current population to accelerate the algorithm's convergence. The systematic experiment results show SA-IMODE has a significant advantage over thirteen state-of-the-art algorithms on benchmark problems.

Semantic Point Cloud Segmentation with Deep-Learning-Based Approaches for the Construction Industry: A Survey

Point cloud learning has recently gained strong attention due to its applications in various fields, like computer vision, robotics, and autonomous driving. Point cloud semantic segmentation (PCSS) enables the automatic extraction of semantic information from 3D point cloud data, which makes it a desirable task for construction-related applications as well. Yet, only a limited number of publications have applied deep-learning-based methods to address point cloud understanding for civil engineering problems, and there is still a lack of comprehensive reviews and evaluations of PCSS methods tailored to such use cases. This paper aims to address this gap by providing a survey of recent advances in deep-learning-based PCSS methods and relating them to the challenges of the construction industry. We introduce its significance for the industry and provide a comprehensive look-up table of publicly available datasets for point cloud understanding, with evaluations based on data scene type, sensors, and point features. We address the problem of class imbalance in 3D data for machine learning, provide a compendium of commonly used evaluation metrics for PCSS, and summarize the most significant deep learning methods developed for PCSS. Finally, we discuss the advantages and disadvantages of the methods for specific industry challenges. Our contribution, to the best of our knowledge, is the first survey paper that comprehensively covers deep-learning-based methods for semantic segmentation tasks tailored to construction applications. This paper serves as a useful reference for prospective research and practitioners seeking to develop more accurate and efficient PCSS methods.

DOA estimation of underwater acoustic co-frequency sources for the coprime vector sensor array

A coprime array with fewer sensors can achieve the same resolution as a uniform linear array. However, when detecting co-frequency targets, there can be prominent false alarms due to overlaps between the main and grating lobes of subarrays. This study proposes a direction-of-arrival (DOA) estimation method to obtain the co-frequency target directions from high grating lobes. The method utilizes joint processing of sound pressure and vibration velocity data from vector hydrophones of a coprime vector hydrophone array and designs joint-cross terms (JCTs) using channel combinations. Based on JCTs, we establish a characteristic data point identification algorithm. The method in this paper can stably and accurately acquire co-frequency target directions from high grating lobes without decoherence operation. Simulation results demonstrate that the proposed algorithm achieves accurate DOA estimation even with reduced signal-to-noise ratio (SNR) and fewer data points. Additionally, a sea experiment confirms the rationality and efficiency of the proposed algorithm, providing new ideas for co-frequency source detection using coprime vector sensor arrays.