Statistical Performance Research Articles

An efficient ‘P1’ algorithm for dual mixed-integer least-squares problems with scalar real-valued parameters

Abstract In this contribution we consider mixed-integer least-squares problems, where the integer ambiguities a ∈ Z n $a\in {\mathbb{Z}}^{n}$ and real-valued parameters b ∈ R p $b\in {\mathbb{R}}^{p}$ are estimated. Both a primal and a dual formulation can be considered, with the latter concerning the ambiguity resolution process taking place into the parameters’ domain. We study the p = 1 case, where an ad hoc ‘P1’ algorithm is introduced, and some geometrical insights are provided. It is demonstrated how the algorithm’s complexity (i.e. number of candidate integer solutions to be evaluated) grows linearly with the ambiguity dimensionality n, differently from the primal formulation where an exponential growth is observed. By means of numerical simulations, here based on Global Navigation Satellite System (GNSS) models, we show the efficiency of this proposed ‘P1’ algorithm, meanwhile also demonstrating its quasi-optimal statistical performance.

Comparative analysis of ensemble learning algorithms in water quality prediction

ABSTRACT Water is an essential resource necessary for the survival of all life forms, yet it is continually at risk of contamination. Accurate water quality prediction is essential for protecting ecosystem health. This study aims to assess the effectiveness of ensemble learning techniques, namely AdaBoost, gradient boosting, XGBoost, CatBoost, and LightGBM, in predicting water quality parameters in the Bara River Basin, Pakistan. Initially, a random forest model was used to determine the input water quality parameters combination for the selected target water quality variable. Then, the ML models were developed for each combination of input parameters and target water quality variables. The ML model's performance was assessed via statistical performance indicators, namely R2, mean squared error, and mean absolute error. The most suitable model was highlighted using compromise programming. The results reveal that the XGBoost and gradient boosting models outperform other algorithms based on statistical indicators, displaying remarkable predictive ability with near-perfect R2 values for HCO3, CO3, and Mg on the XGBoost model and electrical conductivity, SO4, Temp, and Ca on the gradient boosting model. Whereas CatBoost and LightGBM have a more robust performance on some parameters, such as pH and dissolved solids while its performance in other water quality parameters was weak.

The Impact of Socio-Economic and Climate Change on Poverty in Indonesia

Climate change can impact farmers’ incomes as agricultural production still depends on the weather. Currently, the majority of the impoverished rely primarily on agriculture for their income. The connection between poverty and climate change has been extensively studied, but further research is needed in this area. This research was conducted to provide empirical evidence regarding the impact of climate change on poverty using time series data, which has never been done. This research wants to examine the impact of socio-economics (economic growth, agricultural sector growth, inequality, inflation) and climate change on poverty. This research uses time series data from 2007 to 2022. The Central Bureau of Statistics and Climate Change Performance Index (CCPI) reports are the sources of research data. The study results suggest that the government’s performance index in combating inflation, agricultural sector growth, and climate change has a positive impact on poverty. Poverty is negatively affected by the Gini index and economic growth. Government efforts to adaptively address climate change are necessary to prevent worsening impacts on poverty rates. To reduce the risk of crop failure, farmers must also practice practical agricultural management.

Neural network procedures for the cholera disease system with public health mediations

Severe gastrointestinal infections and watery diseases like cholera are still a major worldwide medical concern in the developing nations. A mathematical system contains some necessary dynamics based on the cholera spread to investigate the influence of public health education movements along with treatment and vaccination as control policies in restraining the infection. The cholera disease system with public health mediations divide the density of human population into seven categories based on the status of diseases, who are susceptible, educated, vaccinated, quarantined, infected, treated and removed individuals along with the aquatic bacteria population. The motive of current research is to present the numerical performances of the cholera disease system with public health mediations by using a stochastic computing process based on the Bayesian regularization neural network. A data is constructed by using a conventional Adam scheme that reduces the mean square error by distributing the data into training, validation and testing with some reasonable percentages. Twenty-five neurons, and sigmoid fitness function are used in the stochastic process to solve the model. The accuracy is justified by using comparison of the results, absolute error around 10-06 to 10-08 and some statistical operator performances.

Machine learning approaches to injury risk prediction in sport: a scoping review with evidence synthesis

ObjectiveThis study reviewed the current state of machine learning (ML) research for the prediction of sports-related injuries. It aimed to chart the various approaches used and assess their efficacy, considering...

A saddlepoint approximation for the smoothed periodogram

Poor variance properties of the periodogram often limit its practical applicability to a wide range of modern spectral estimation and detection applications. The smoothed periodogram, a refined periodogram-based method, is one such nonparametric approach to reducing variance. Neighboring spectral samples are averaged across a spectral window, as opposed to the more common temporal or lag window. Tapered spectral windows and other modifications needed to address the time-bandwidth product and resolution-variance trade-offs complicate the statistical analysis, making it difficult to quantify statistical performance. In addition, approximate distributions for the smoothed periodogram require a priori normalization along with simplifying assumptions to yield computationally tractable results. Here, under mild asymptotic conditions, the distribution derived prior to normalization is shown to be computationally intractable in most cases. First-order statistical approximations are computationally stable but result in sizeable inaccuracies, particularly in the tails. We use a saddlepoint approximation, a second-order asymptotic method, that allows for accurate statistical characterization but is also numerically stable. Monte Carlo simulations are used to validate the results and to illustrate the robustness of the approach. Finally, its utility is demonstrated on a real-world dataset relevant to the side-channel and hardware cybersecurity communities.

Assessment of the Statistical Performance of Chemical Transport Model Studies in India

Assessment of the Statistical Performance of Chemical Transport Model Studies in India

Modeling farmer behavior in adopting agroecological practices using AI tools in shea butter agroforestry systems

This study examines the performance of machine learning algorithms for identifying importance features for agroecological practices adoption in shea agroforestry systems. Primary data were collected from 272 representative and randomly selected farmers in two regions of northern Benin. Four machine learning algorithms (Naïve Bayes, Neural Network, Support Vector Machine and Bagging Decision Trees) were compared using four statistical performance metrics: accuracy, balanced accuracy, recall, and the area under the receiver operating characteristic curve (AUC), as well as calibration plots. The results indicated that the Naïve Bayes model performed best for predicting animal parking practices, and biopesticide while the Support Vector Machine outperformed other models in predicting cultural associations, improved seeds, and organic fertilizers adoption. Feature importance analysis revealed that the most significant feature influencing the adoption of animal parking and organic fertilizers practices was the farmers' region of origin, with the highest importance value (100%). When considering features with at least 50% importance, main occupation and means of transport for manure were also relevant for the adoption of animal parking practice. The farmers' ethnic group, village, and age class were the most influential characteristics for biopesticides, cultural association, and improved seeds practices, respectively, each with a 100% importance value. For improved seeds practice, farm assets and secondary occupation were also relevant when considering features with at least 50% importance. Therefore, providing comprehensive training on agroecological practices such as animal parking, crop diversification, biopesticide management, organic fertilization, and improved seeds use to farmers in Northern Benin, particularly those managing shea agrosystems, could contribute to the sustainable development of this vital resource and enhance food security.

Daily Estimates of Global Radiation in the Brazilian Amazon from Simplified Models

Solar radiation is an element and a meteorological factor that is present in several processes, such as evapotranspiration, photosynthesis, and energy generation, among others. However, in some regions, there is a limitation in surface data measurements. In this study, 87 empirical models were evaluated to estimate global radiation (Hg) in the Brazilian Amazon biome; these models were divided into five groups according to the input variables, such as insolation (group I), air temperature (group II), relative humidity (group III), astronomical variables (group IV), and hybrid models (group V). The estimates were evaluated by their significance (t-test) and then according to the statistical metrics of the models’ performance (R2, MBE, RMSE, and d). The group V model Hg/H0=a+bln∆T+cS/S0d presented the best statistical performance in all the evaluated indicators, followed by the group I model Hg/H0=a+bS/S0c, and then the group II model Hg/H0=a+b∆T+c∆T0.25+d∆T0.5+eTmed/H0. The group III models presented a low statistical performance, and the group IV models were not significant (NS) in all the evaluated meteorological stations. In general, it was found that the simplified or hybrid models based on insolation and air temperature were efficient in estimating Hg in the Brazilian Amazon biome.

Electricity demand uncertainty modeling with Temporal Convolution Neural Network models

This work presents a Temporal Convolution Network (TCN) model for half-hourly, three-hourly and daily-time step to predict electricity demand (G) with associated uncertainties for sites in Southeast Queensland Australia. In addition to multi-step predictions, the TCN model is applied for probabilistic predictions of G where the aleatoric and epistemic uncertainties are quantified using maximum likelihood and Monte Carlo Dropout methodologies. The benchmarks of TCN model include an attention-based, bi-directional, gated recurrent unit, seq2seq, encoder–decoder, recurrent neural networks and natural gradient boosting models. The testing results show that the proposed TCN model attains the lowest relative root mean square error of 5.336-7.547% compared with significantly larger errors for all benchmark models. In respect to the 95% confidence interval using the Diebold–Mariano test statistic and key performance metrics, the proposed TCN model is better than benchmark models, capturing a lower value of total uncertainty, as well as the aleatoric and epistemic uncertainty. The root mean square error and total uncertainty registered for all of the forecast horizons shows that the benchmark models registered relatively larger errors arising from the epistemic uncertainty in predicted electricity demand. The results obtained for TCN, measured by the quality of prediction intervals representing an interval with upper and lower bound errors, registered a greater reliability factor as this model was likely to produce prediction interval that were higher than benchmark models at all prediction intervals. These results demonstrate the effectiveness of TCN approach in electricity demand modelling, and therefore advocates its usefulness in now-casting and forecasting systems.

Enhancing security and randomness in cryptography and non-cryptographic applications with ORNA algorithm

Abstract In this paper, we advocate for using PRNGs to generate long sequences of statistically random numbers, also known as pseudorandom numbers. It has several practical applications, including cryptography, weather-based stimulation, procedural environment noise generation in games and casinos, and scientific simulations of real-world problems, like city traffic systems that use it to stimulate cars. One distinctive feature of our approach is the PRNG algorithm we propose; this algorithm has successfully produced statistically sound long-term sequences. The properties of the PRNG are assessed using several statistical methods. Our work presents the Oscillating Random Number Algorithm (ORNA), a new kind of random number generator, and examines both the statistical tests proposed by NIST and the results that ORNA has generated. The following also describes the several preceding algorithms developed, including the Mersenne Twister Algorithm (MT19937) and the Permuted Congruential Algorithm (PCG128), and compares ORNA’s graphical efficiency to theirs. In this work, we compared ORNA to MT19937 and PCG128, and found that ORNA excels according to all 16 statistical tests. ORNA has three primary benefits, namely, superior statistical performance, quicker code execution, and simpler code implementation due to the algorithm’s reduced complexity.

Blockchain-inspired intelligent framework for logistic theft control

The smart logistics industry utilizes advanced software and hardware technologies to enhance efficient transmission. By integrating smart components, it identifies vulnerabilities within the logistics sector, making it more susceptible to physical attacks aimed at theft and control. The main goal is to propose an effective logistics monitoring system that automates theft prevention. Specifically, the suggested model analyzes logistics transmission patterns through secure surveillance enabled by IoT-based blockchain technology. Additionally, a bi-directional convolutional neural network is employed to evaluate real-time theft vulnerabilities, aiding optimal decision-making. The proposed method has been shown to provide accurate real-time analysis of risky behaviors. Experimental simulations indicate that the proposed solution significantly improves logistics monitoring. The system’s performance is assessed using various statistical metrics, including latency rate (7.44 s), a data processing cost (O((n−1)logn)), and model training and testing results (precision (94.60%), recall (95.67%), and F-Measure (96.64%)), statistical performance (error reduction (48%)) and reliability (94.48%).

Memetic ant colony optimization for multi-constrained cognitive diagnostic test construction.

Cognitive diagnostic tests (CDTs) assess cognitive skills at a more granular level, providing detailed insights into the mastery profile of test-takers. Traditional algorithms for constructing CDTs have partially addressed these challenges, focusing on a limited number of constraints. This paper intends to utilize a meta-heuristic algorithm to produce high-quality tests and handle more constraints simultaneously. This paper presents a memetic ant colony optimization (MACO) algorithm for constructing CDTs while considering multiple constraints. The MACO method utilizes pheromone trails to represent successful test constructions from the past. Additionally, it innovatively integrates item quality and constraint adherence into heuristic information to manage multiple constraints simultaneously. The method evaluates the assembled tests based on the diagnosis index and constraint satisfaction. Another innovation of MACO is the incorporation of a local search strategy to further enhance diagnostic accuracy by partially optimizing item selection. The optimal local search parameter settings are explored through a parameter investigation. A series of simulation experiments validate the effectiveness of MACO under various conditions. The results demonstrate the great ability of meta-heuristic algorithms to handle multiple constraints and achieve high statistical performance. MACO exhibited superior performance in generating high-quality CDTs while meeting multiple constraints, particularly for mixed and low discrimination item banks. It achieved faster convergence than the ant colony optimization in most scenarios. MACO provides an effective solution for multi-constrained CDT construction, especially for shorter tests and item banks with mixed or lower discrimination. The experimental results also suggest that the suitability of different optimization approaches may depend on specific test conditions, such as the characteristics of the item bank and the length of the test.

Minus the Error: Estimating d N / d S and Testing for Natural Selection in the Presence of Residual Alignment Errors.

Errors in multiple sequence alignments (MSAs) are known to bias many comparative evolutionary methods. In the context of natural selection analyses, specifically codon evolutionary models, excessive rates of false positives result. A characteristic signature of error-driven findings is unrealistically high estimates of dN/dS (e.g., >100), affecting only a small fraction (e.g., ∼0.1%) of the alignment. Despite the widespread use of codon models to assess alignment quality, their potential for error correction remains unexplored. We present BUSTED-E: a novel method designed to detect positive selection while concurrently identifying alignment errors. This method is a straightforward adaptation of the BUSTED flexible branch-site random effects model used to fit distributions of dN/dS, with an important modification: it integrates an "error-sink" component representing an abiological evolutionary regime (dN/dS > 100), and provides the option for masking errors in the MSA for downstream analyses. Statistical performance of BUSTED-E on data simulated without errors shows that there is a small loss of power, which can be mitigated by model averaged inference. Using four published empirical datasets, we show BUSTED-E reduces unrealistic rates of positive selection detection, often by an order of magnitude, and improves biological interpretability of results. BUSTED-E also detects errors that are largely distinct from other popular alignment cleaning tools (HMMCleaner and BMGE). Overall, BUSTED-E is a robust and scalable solution for improving the accuracy of evolutionary analyses in the presence of residual alignment errors, contributing to a more nuanced understanding of natural selection and adaptive evolution.

Performance of an AI-powered visualization software platform for precision surgery in breast cancer patients

Surgery remains the primary treatment modality in the management of early-stage invasive breast cancer. Artificial intelligence (AI)-powered visualization platforms offer the compelling potential to aid surgeons in evaluating the tumor’s location and morphology within the breast and accordingly optimize their surgical approach. We sought to validate an AI platform that employs dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) to render three-dimensional (3D) representations of the tumor and 5 additional chest tissues, offering clear visualizations as well as functionalities for quantifying tumor morphology, tumor-to-landmark structure distances, excision volumes, and approximate surgical margins. This retrospective study assessed the visualization platform’s performance on 100 cases with ground-truth labels vetted by 2 breast-specialized radiologists. We assessed features including automatic AI-generated clinical metrics (e.g., tumor dimensions) as well as visualization tools including convex hulls at desired margins around the tumor to help visualize lumpectomy volume. The statistical performance of the platform’s automated features was robust and within the range of inter-radiologist variability. These detailed 3D tumor and surrounding multi-tissue depictions offer both qualitative and quantitative comprehension of cancer topology and may aid in formulating an optimal surgical approach for breast cancer treatment. We further establish the framework for broader data integration into the platform to enhance precision cancer care.

Forecasting Blue and Green Water Footprint of Wheat Based on Single, Hybrid, and Stacking Ensemble Machine Learning Algorithms Under Diverse Agro-Climatic Conditions in Nile Delta, Egypt

The aim of this research is to develop and compare single, hybrid, and stacking ensemble machine learning models under spatial and temporal climate variations in the Nile Delta regarding the estimation of the blue and green water footprint (BWFP and GWFP) for wheat. Thus, four single machine learning models (XGB, RF, LASSO, and CatBoost) and eight hybrid machine learning models (XGB-RF, XGB-LASSO, XGB-CatBoost, RF-LASSO, CatBoost-LASSO, CatBoost-RF, XGB-RF-LASSO, and XGB-CatBoost-LASSO) were used, along with stacking ensembles, with five scenarios including climate and crop parameters and remote sensing-based indices. The highest R2 value for predicting wheat BWFP was achieved with XGB-LASSO under scenario 4 at 100%, while the minimum was 0.16 with LASSO under scenario 3 (remote sensing indices). To predict wheat GWFP, the highest R2 value of 100% was achieved with RF-LASSO across scenario 1 (all parameters), scenario 2 (climate parameters), scenario 4 (Peeff, Tmax, Tmin, and SA), and scenario 5 (Peeff and Tmax). The lowest value was recorded with LASSO and scenario 3. The use of individual and hybrid machine learning models showed high efficiency in predicting the blue and green water footprint of wheat, with high ratings according to statistical performance standards. However, the hybrid programs, whether binary or triple, outperformed both the single models and stacking ensemble.

Monitoring of zero-inflated COM-Poisson processes

This paper investigates control charts for count data with a significant number of zeros. The proposed models are generalized to handle data exhibiting different types of dispersion, i.e. equidispersion, overdispersion, and underdispersion. To this effect, the zero-inflated Conway–Maxwell Poisson (ZICMP) distribution is employed. A Shewhart and an exponentially weighted moving average (EWMA) control chart are developed, referred to as ZICMP-Shewhart and ZICMP-EWMA charts, respectively. The ZICMP distribution incorporates various distributions, including the Conway–Maxwell–Poisson (COM-Poisson) and zero-inflated Poisson (ZIP), Geometric, and Bernoulli distributions, as special cases. Consequently, the flexibility and versatility of the ZICMP distribution enhance the applicability of the proposed control charts, thereby providing practitioners with an adaptable tool suitable for various scenarios.The control charts are examined for their ability to detect upward shifts in the process mean level, and their statistical performance is evaluated in terms of the average run-length (ARL). Through a simulation study, we demonstrate that the ZICMP-Shewhart chart is more effective for monitoring over-dispersed data, while the ZICMP-EWMA chart is more suitable for under-dispersed data. Finally, we provide two real-life examples to illustrate the applicability of the proposed charts.

Hotelling T2 control chart based on minimum vector variance for monitoring high‐dimensional correlated multivariate process

AbstractMultivariate control charts are practical tools that simultaneously monitor several correlated quality characteristics in a process. Monitoring high‐dimensional data structures is challenging because, in most cases, the process sample size for monitoring parameters is greater than the number of process characteristics. Many researchers have used the multivariate Hotelling's T2 chart to monitor high‐dimensional data using the maximum‐likelihood methods (MLM) to estimate the covariance matrices. However, the multivariate Hotelling's T2chart based on MLM suffers from low statistical performance. In this paper, we proposed a multivariate Hotelling's T2 chart based on the minimum vector variance (MVV) and some regularized methods for monitoring high‐dimensional data structures. The performance of the proposed chart is evaluated in terms of the average run length (ARL). The results reveal the superiority of the proposed MVV Hotelling's T2 chart over the existing Hotelling's T2 charts for high‐dimensional correlated processes.

The Integration of Internet of Things and Machine Learning for Energy Prediction of Wind Turbines

Wind power has emerged as a crucial substitute for conventional fossil fuels. The combination of advanced technologies such as the internet of things (IoT) and machine learning (ML) has given rise to a new generation of energy systems that are intelligent, reliable, and efficient. The wind energy sector utilizes IoT devices to gather vital data, subsequently converting them into practical insights. The aforementioned information aids among others in the enhancement of wind turbine efficiency, precise anticipation of energy production, optimization of maintenance approaches, and detection of potential risks. In this context, the main goal of this work is to combine the IoT with ML in the wind energy sector by processing weather data acquired from sensors to predict wind power generation. To this end, three different regression models are evaluated. The models under comparison include Linear Regression, Random Forest, and Lasso Regression, which were evaluated using metrics such as coefficient of determination (R²), adjusted R², mean squared error (MSE), root mean squared error (RMSE), and mean absolute error (MAE). Moreover, the Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) were taken into consideration as well. After examining a dataset from IoT devices that included weather data, the models provided substantial insights regarding their capabilities and responses to preprocessing, as well as each model’s reaction in terms of statistical performance deviation indicators. Ultimately, the data analysis and the results from metrics and criteria show that Random Forest regression is more suitable for weather condition datasets than the other two regression models. Both the advantages and shortcomings of the three regression models indicate that their integration with IoT devices will facilitate successful energy prediction.

Evaluating inverse modeling methods for measurement and verification of chiller energy efficiency measures

A commonly budget and time constrained yet crucial measurement and verification process for chiller energy efficiency measures often requires rarely trended chiller performance data for the pre-retrofit or post-retrofit chiller necessitating chiller modeling for year-round performance prediction. This study evaluates chiller inverse modeling methods for measurement and verification applications. Variable speed drive-controlled centrifugal chillers operating in a hot and humid climate are used as case studies. Two scenarios are explored: one where full-range metered chiller data are available and another with limited data that requires a short-term metering process. The biquadratic black-box and Gordon-Ng with a variable entropy term models perform exceptionally well when full-range chiller performance data is available, displaying a coefficient of variation of approximately 5 % for both training and test datasets. However, in situations that use short-term metered data not covering the full range, the fundamental and Foliaco reformulated Gordon-Ng models outperform other models. These models show minimal degradation in average statistical performance when trained with one-month metering data instead of four-month data, indicating their potential to capture the year-round chiller performance variation with short-term metered data. Furthermore, the optimal metering period for an accurate modeling process is identified as one containing data from at least one shoulder month like April, May, or October for a hot and humid climate. These findings provide valuable guidance for practitioners involved in chiller efficiency measure assessments, emphasizing the significance of proper model selection and an appropriate metering period for a reliable measurement and verification process.