Magnitude Of Relative Error Research Articles

Theoretical analysis of water-droplet condensation and evaporation in prescribed environment with radiation

Abstract A quasi-explicit algebraic solution of the problem of quasi-steady water droplet heat- and mass-transfer with condensation or evaporation in a prescribed environment including radiation has been obtained. The solution is based on the mass-fraction form of Fick’s law of species diffusive mass transport, which allows accounting for variation in gas density through the diffusion layer surrounding the droplet. A new term, the average vapor supersaturation through the diffusion layer, has been included that allows the model to be extended to environmental conditions significantly outside those for which it is theoretically derived (dilute or near-saturation), even to conditions of relatively high temperature, low relative humidity, and low pressure. The latter two have atmospheric significance because evaporative cooling can induce significant decrease in droplet temperature and possibly induce freezing. Basic modeling assumptions are confirmed by comparison with time-dependent temperature and evaporation-rate measurements of a laboratory laser-heated droplet. Accuracy of the new model is assessed by comparison with exact equations. For temperatures between −30 and 0°C and negligible radiation, relative-error magnitudes in droplet temperature and evaporation rate are less than 5%, even for relative humidity down to 10% and pressure down to 60 kPa. This represents an improvement in accuracy over the conventional model, in which property variations of density and supersaturation through the diffusion layer are not taken into account. Differences in predicted evaporation rates between the mass-fraction and the more common partial-density Fick’s laws, which can be significant (>10% above 70% relative humidity) in coupled, implicit equations, are shown to be made negligibly small (<0.5%) by linearizing about the saturation state.

Optimal wideband digital fractional-order differentiators using gradient based optimizer

In this paper, we propose a novel optimization approach to designing wideband infinite impulse response (IIR) digital fractional order differentiators (DFODs) with improved accuracy at low frequency bands. In the new method, the objective function is formulated as an optimization problem with two tuning parameters to control the error distribution over frequencies. The gradient based optimizer (GBO) is effectively employed on the proposed objective function. A wide range of design examples are presented to illustrate the effectiveness of the proposed approach. The proposed approximations are compared to those of recent literature in terms magnitude, phase, and group delay errors. The result reveal that our method can attain approximations have a favorable low frequency performance (with about 60% of relative magnitude error reduction) and maintain a comparable accuracy at most of the Nyquist band to those of the existing ones. Thus, our approximations can be attractive for low frequency applications.

Data-driven prediction of future melt pool from built parts during metal additive manufacturing

Smart manufacturing in metal additive manufacturing (MAM) relies on real-time process optimization through the prediction of unbuilt parts using data from built parts. Melt pool dynamics are intimately associated with the development of various microstructures during the MAM. In this paper, we demonstrate the idea by establishing a machine learning (ML) model to predict the melt pool of future parts, based on the experimental data from the National Institute of Standards and Technology (NIST) where 80 % is used for the training (built parts) and 20 % for testing (future parts). The ML model integrates both data denoising and predictive modeling. First, a convolutional neural network (CNN) is used to process a large dataset of raw melt pool images, enabling the automatic removal of noises (e.g., splash and plume) and thereof extraction of high-quality melt pool data. Following that, a novel data-driven melt pool model based on multi-layer perceptron (MLP) is trained by incorporating raw, long scanning history as input features, which best accounts for the effects of printing history (e.g., intertrack heating) on melt pool development. It takes complete advantage of MLP in handling high-dimensional regression problems in conjunction with a large dataset. Upon testing under various manufacturing conditions, the average relative error magnitude (AREM) of predicting melt pool size drops to 2.8 %, compared to 14.8 % of the prior art – the Neighboring Effect Modeling Method model (NBEM). This research thus represents a significant step towards reliable melt-pool-guided AM process optimization for smart manufacturing, enabled by advanced, flexible, and maximum use of ML techniques.

Software cost and effort estimation using dragonfly whale optimized multilayer perceptron neural network

This paper proposed a Constructive Rapid Application Development Model based Software Cost Estimation Technique (CORADMO based SCET) for accurate software cost estimation. The data requirements, cost drivers, constraints and priorities are given as an input to the Fuzzy Inference System (FIS). The processed output such as effort, time and cost for nominal plan, shortest schedule plan, and least cost plan are computed in the Fuzzy Inference System (FIS). Further to reduce the effort time and cost means, the output optimized by Dragon fly Whale Optimization (DWO)which provides the best estimated effort, time and cost as an output for software development. The proposed CORADMO based SCET model is evaluated by NASA 93 dataset using MATLAB. The performance of the CORADMO based SCET approach is assessed in terms of Mean Magnitude of Relative Error, Pred (25%), and Magnitude of Relative Error attains the values of 98.77%, 92.55%, and 93.45% respectively. Finally, the CORADMO based SCET model justifies the suitability of Dragon fly Whale Optimization with the proposed fuzzy logic.

Software cost estimation predication using a convolutional neural network and particle swarm optimization algorithm

Over the past decades, the software industry has expanded to include all industries. Since stakeholders tend to use it to get their work done, software houses seek to estimate the cost of the software, which includes calculating the effort, time, and resources required. Although many researchers have worked to estimate it, the prediction accuracy results are still inaccurate and unstable. Estimating it requires a lot of effort. Therefore, there is an urgent need for modern techniques that contribute to cost estimation. This paper seeks to present a model based on deep learning and machine learning techniques by combining convolutional neural networks (CNN) and the particle swarm algorithm (PSO) in the context of time series forecasting, which enables feature extraction and automatic tuning of hyperparameters, which reduces the manual effort of selecting parameters and contributes to fine-tuning. The use of PSO also enhances the robustness and generalization ability of the CNN model and its iterative nature allows for efficient discovery of hyperparameter similarity. The model was trained and tested on 13 different benchmark datasets and evaluated through six metrics: mean absolute error (MAE), mean square error (MSE), mean magnitude relative error (MMRE), root mean square error (RMSE), median magnitude relative error (MdMRE), and prediction accuracy (PRED). Comparative results reveal that the performance of the proposed model is better than other methods for all datasets and evaluation criteria. The results were very promising for predicting software cost estimation.

A Hybrid Artificial Bee Colony and Artificial Fish Swarm Algorithms for Software Cost Estimation

Software cost estimation (SCE), estimating the cost and time required for software development, plays a highly significant role in managing software projects. A somewhat accurate SCE is necessary for a software project to be successful. It allows effective control of construction time and cost. In the past few decades, various models have been presented to evaluate software projects, including mathematical models and machine learning algorithms. In this paper, a new model based on the hybrid of the artificial fish swarm algorithm (AFSA) and the artificial bee colony (ABC) algorithm is presented for SCE. The initial population of AFSA, which includes the values of the effort factors, is generated using the ABC algorithm. ABC algorithm is used to solve the problems of the AFSA algorithm such as population diversity and getting stuck in a local optimum. ABC algorithm achieves the best solutions using observer and scout bees. The evaluation of the combined method has been implemented on eight different data sets and evaluated based on eight different criteria such as mean magnitude of relative error and PRED (0.25). The proposed method is more error-free than current SCE methods, according to the results. The error value of the proposed method is lower on NASA60, NASA63, and NASA93 datasets.

A NONLINEAR REGRESSION MODEL FOR EARLY LOC ESTIMATION OF OPEN-SOURCE KOTLIN-BASED APPLICATIONS

Context. The early lines of code (LOC) estimation in software projects holds significant importance, as it directly influences the prediction of development effort, covering a spectrum of different programming languages, and open-source Kotlin-based applications in particular. The object of the study is the process of early LOC estimation of open-source Kotlin-based apps. The subject of the study is the nonlinear regression models for early LOC estimation of open-source Kotlin-based apps. Objective. The goal of the work is to build the nonlinear regression model with three predictors for early LOC estimation of open-source Kotlin-based apps based on the Box-Cox four-variate normalizing transformation to increase the confidence in early LOC estimation of these apps. Method. For early LOC estimation in open-source Kotlin-based apps, the model, confidence, and prediction intervals of nonlinear regression were constructed using the Box-Cox four-variate normalizing transformation and specialized techniques. These techniques, relying on multiple nonlinear regression analyses incorporating multivariate normalizing transformations, account for the dependencies between variables in non-Gaussian data scenarios. As a result, this method tends to reduce the mean magnitude of relative error (MMRE) and narrow confidence and prediction intervals compared to models utilizing univariate normalizing transformations. Results. An analysis has been carried out to compare the constructed model with nonlinear regression models employing decimal logarithm and Box-Cox univariate transformation. Conclusions. The nonlinear regression model with three predictors for early LOC estimation of open-source Kotlin-based apps is constructed using the Box-Cox four-variate transformation. Compared to the other nonlinear regression models, this model demonstrates a larger multiple coefficient of determination, a smaller value of the MMRE, and narrower confidence and prediction intervals. The prospects for further research may include the application of other data sets to construct the nonlinear regression model for early LOC estimation of open-source Kotlin-based apps for other restrictions on predictors.

Activity-based function point complexity of use case diagrams for software effort estimation

This study proposes a Function Point Analysis (FPA) based software development effort estimation methodology integrated with Use Case Diagrams. These methods include identifying actor activities, classifying those activities into FPA categories, and calculating Unadjusted Function Points (UFP). Followed by the calculation of Technical Complexity Factors (TCF) and Adjusted Function Points (AFP), this study aims to produce more accurate man-hours estimates. Results show a UFP of 162 TCF of 11, AFP of 123.12, and an estimated effort of 1846.8 hours worked, while the actual effort is 1228 hours. Evaluation of estimates using the metrics Mean Magnitude of Relative Error (MMER) 0.34, Mean Magnitude of Relative Error (MMRE) 0.50, Mean Absolute Error (MAE) 618.80, Mean Balanced Relative Error (MBRE) 0.50, Mean Inverse Balanced Relative Error (MIBRE) 0.34, and Root Mean Squared Error (RMSE) 618.80, showed sufficient precision despite the overestimation. The study suggests the need for adjustments in TCF calculations and considering development environment factors in more detail to improve estimation accuracy. These findings are essential in improving the precision of effort estimation methodologies in software development, particularly in projects that use Use Case Diagrams as the primary framework.

MFLion‐DMN: Mayfly Lion‐optimized deep maxout network for prediction of software development effort

AbstractThe precise evaluation of predicting software development effort is a serious process in software engineering. The underestimates require more time, which seems to be the negotiation of complete efficient design and complete software testing. Thus, this paper devises a novel model for inspecting the effort taken for the design of software. The Mayfly Lion Optimization (MFLion) algorithm is devised, which ensembles the Mayfly Algorithm (MA) with the Lion Optimization Algorithm (LOA) for precise estimation. The deep maxout network (DMN) is adapted wherein the different weights are produced out of which the most suitable is infused considering the MFLion during model training. The optimal features are selected using recursive feature elimination (RFE) wherein the best features are selected and the remaining irrelevant features are eliminated from the list. The proposed MFLion obtained better performance with the smallest mean magnitude of relative error (MMRE) of 5.909 and the smallest root mean square error (RMSE) of 75.505, respectively. Each technique is produced using a separate database generated using the Promise software engineering repository. The outcomes produced from the assessment of the accuracies of models suggested that the MFLion‐DMN is a substitute to forecast software effort, which is extensively devised in engineering platforms.

Numerical Modeling of Tidal Phenomena of the Penzhinskaya Bay

In this study, the Delft3D-Flow hydrodynamic model was used to determine the characteristics of the main tidal waves and currents in the Penzhina Bay (Sea of Okhotsk, Russia). The initial and boundary conditions are set according to the TPXO9 model data, the salinity and water temperature for each model layer are set based on the reanalysis data of the HYCOM ocean general circulation model, with a time step of 3 h. E-uropean Center for Medium-Range Weather Forecasts ECMWF (European Center for Medium-Range Weather Forecasts) reanalysis ERA-5 was used as meteoforcing. Modeling was performed for the ice-free period from May to September 2005. Maps of cotidal lines, tidal ellipses, and reversibility coefficient for the main tide waves: semidiurnal wave M2 and diurnal wave K1 were constructed. The model was verified by comparing the published and calculated harmonic components for 9 sea level monitoring points located in the Penzhina Bay. Conformance evaluation generally shows a high degree of consistency between model and reference data. The values of the coefficient of determination R2 between the series formed by the model and published harmonic components are in the range of 0.96–0.99. According to the magnitude of relative errors, the simulation results are divided into two consistency categories – high (1.48–2.14%) and satisfactory (2.93–4.27%). Spatial patterns of distribution for the values of relative errors were not found. A certain inconsistency in the results is presumably due to the time discretization of the observation time support and the short sea level monitoring data series used to calculate the published harmonic components at the Penzhina Bay.

AGILE ESTIMATIONS: EMPIRICAL STUDY USING WEIGHTED COMPLEXITY FACTORS

Software time and cost estimates are always challenging due to the quickly changing software process. Especially Agile project estimations are complex, and the story points differ from person to person. Due to the person-to-person variations, project managers have difficulty arriving at the project cost for development projects with different sizes or complexity of systems/software. This paper proposes a new estimation model named the WEighted COmplexity Factor Estimation (WECOFE) Model to reduce the complexities of Agile story point estimations. Weighted complexities factors identify Low, Moderate, and High Project Level Factors and Story Level Factors. For each level, multiple factors were defined. Derived Weighted complexity based on Project, Requirement, and Resource Level factors. Each story's point estimation is explained based on weighted complexity defined and assigned factors values. Conducted study on various industry projects, showing that the WECOFE approach is promising by observing the Magnitude of Relative Error (MRE) and Mean Magnitude of Relative Error (MMRE). Also, performed a t-test for one sprint project with 51 stories, and paired sample means are equal for WECOFE with actual story points spent by the team.

Incorporating statistical and machine learning techniques into the optimization of correction factors for software development effort estimation

AbstractAccurate effort estimation is necessary for efficient management of software development projects, as it relates to human resource management. Ensemble methods, which employ multiple statistical and machine learning techniques, are more robust, reliable, and accurate effort estimation techniques. This study develops a stacking ensemble model based on optimization correction factors by integrating seven statistical and machine learning techniques (K‐nearest neighbor, random forest, support vector regression, multilayer perception, gradient boosting, linear regression, and decision tree). The grid search optimization method is used to obtain valid search ranges and optimal configuration values, allowing more accurate estimation. We conducted experiments to compare the proposed method with related methods, such as use case points‐based single methods, optimization correction factors‐based single methods, and ensemble methods. The estimation accuracies of the methods were evaluated using statistical tests and unbiased performance measures on a total of four datasets, thus demonstrating the effectiveness of the proposed method more clearly. The proposed method successfully maintained its estimation accuracy across the four experimental datasets and gave the best results in terms of the sum of squares errors, mean absolute error, root mean square error, mean balance relative error, mean inverted balance relative error, median of magnitude of relative error, and percentage of prediction (0.25). The p‐value for the t‐test showed that the proposed method is statistically superior to other methods in terms of estimation accuracy. The results show that the proposed method is a comprehensive approach for improving estimation accuracy and minimizing project risks in the early stages of software development.

DESIGN OF FIRST ORDER DIFFERENTIATOR WITH PARALLEL ALL-PASS STRUCTURE

In this paper a new method for design of the first order differentiator is presented. The proposed differentiator consists of two parallel branches, i.e. direct path and IIR all-pass filter. The described design method allows one to obtain solution with minimum mean relative error at the desired region by controlling the ratio of phase response extremes. A small relative magnitude error, as well as a low phase error, at low frequencies is condition for good time domain behaviour. The obtained differentiator can be realized by means of only two multipliers, hence being a good choice for real time applications. The proposed solution provides a lower magnitude error than several known differentiators with similar phase error.

Deep learning based one step and multi‐steps ahead forecasting blood glucose level

AbstractEnabling diabetic patients to predict their Blood Glucose Levels (BGL) is a crucial aspect of managing their metabolic condition, as it allows them to take appropriate measures to avoid hypo or hyperglycemia. Machine Learning (ML) and Deep Learning (DL) techniques have made this possible, and this paper evaluates and compares the performance of five distinct ML/DL models including: Convolutional Neural Network (CNN), Long Short Term Memory (LSTM), Support Vector Regression (SVR), Gated Reccurent Unit (GRU) and Deep Belief Network (DBN) for forecasting BGL, by applying two different forecasting methods, namely One Step Ahead (OSF) and Multi‐Step Ahead (MSF) comprising five different variants. The performance is evaluated based on four metrics: Mean Absolute Error (MAE), Mean Magnitude Relative Error (MMRE), Root Mean Square Error (RMSE) and Predictive Level (PRED). Additionally, the statistical significance of the regressors was evaluated using the Scott‐Knott (SK) test, while the Borda Count (BC) voting system was employed to rank them. The results indicate that the best performance was achieved with OSF using GRU. Furthermore, the effectiveness of an MSF strategy depends on the ML/DL technique used, and the best combinations were DBN with DirRec, DBN with Recursive, SVR with Recursive and SVR with DirRec. Additionally, DirRec was found to be the best strategy, as it consistently ranked first regardless of the ML/DL technique used.

Transient thermal models of induction machines under Inter‐turn short‐circuit fault conditions

AbstractInduction machines are the working horses in a lot of industries. Inter‐turn short‐circuit (ITSC) fault is one of the most common failure modes taking place in them. It can generate large fault currents that lead to a local temperature rise in the faulty stator slot, which deteriorates the working performance of the machine or even cascades to a complete machine breakdown. This article focuses on developing transient thermal models for an induction machine under ITSC faults. The aim of these thermal models is to understand the thermal behaviour of the induction machine at different ITSC fault propagation stages. The first thermal model is based on finite element method (FEM) simulation that includes a physics‐based model for the thermal contact resistance. The second thermal model is a lumped parameter thermal network that models the stator as inter‐connected slot‐number parts. They are both capable of modelling the characteristic of non‐uniform temperature along the circumferential direction for the induction machine operating under ITSC fault conditions. After area‐weighted averaging the results of the FEM simulation, it is found they are quantitatively consistent, with an average relative error magnitude of 5%, as well.

Identification of multiple radioisotopes through convolutional neural networks trained on 2-D transformed gamma spectral data from CsI(Tl) spectrometer

A study is conducted to improve the performance of Convolutional Neural Networks (CNNs)-based radioisotope identification (RIID). This study introduces a novel method to transform the dimensional structure of a radioisotope (RI) mixture spectra as training inputs, for the purpose of extending the receptive field and improving RIID performance of the model. The one-dimensional (1-D) composite spectra of radioisotopic mixtures as a dataset to be used as train and test data for a CNNs model are generated by linearly combining the normalized element gamma-ray spectra of single RIs obtained experimentally with the developed CsI(Tl) spectrometer. One more dataset for training and testing another a CNNs model is produced by reshaping the same 1-D composite spectra into two-dimensional (2-D) image data with multiple columns and rows. After the structure and hyper-parameters of the model are determined, each model is trained and tested with each of the two prepared datasets to simultaneously predict the probabilities as the relative count distributions for eight RI classes. The RIID performance of the CNNs model trained with the reshaped 2-D data is compared with that of the model trained with the 1-D data. The performance of each trained model is evaluated by mean magnitude of relative error (MMRE) as performance evaluation metrics. The results show that the model trained with 2-D inputs outperforms the model trained with 1-D inputs and effectively identifies unknown multiple radioisotopes. The proposed method to transform 1-D data into 2-D data represents a promising approach for the CNNs-based RIID of a scintillator spectrometer, especially when multiple peaks or RIs need to be resolved.

Effort Estimation in Agile Software Development: A Systematic Map Study

Introduction − Making effort estimation as accurate and suitable for software development projects becomes a fundamental stage to favor its success, which is a difficult task, since the application of these techniques in constant changing agile development projects raises the need to evaluate different methods frequently. Objectives− The objective of this study is to provide a state of the art on techniques of effort estimation in agile software development (ASD), performance evaluation and the drawbacks that arise in its application. Method− A systematic mapping was developed involving the creation of research questions to provide a layout of this study, analysis of related words for the implementation of a search query to obtain related studies, application of exclusion, inclusion, and quality criteria to filter nonrelated studies and finally the organization and extraction of the necessary information from each study. Results− 25 studies were selected; the main findings are: the most applied estimation techniques in agile contexts are: Estimation of Story Points (SP) followed by Planning Poker (PP) and Expert Judgment (EJ). The most frequent solutions supported in computational techniques such as: Naive Bayes, Regression Algorithms and Hybrid System; also, the performance evaluation measures Mean Magnitude of Relative Error (MMRE), Prediction Assessment (PRED) and Mean Absolute Error (MAE) have been found to be the most commonly used. Additionally, parameters such as feasibility, experience, and the delivery of expert knowledge, as well as the constant particularity and lack of data in the process of creating models to be applied to a limited number of environments are the challenges that arise the most when estimating software in agile software development (ASD) Conclusions− It has been found there is an increase in the number of articles that address effort estimation in agile development, however, it becomes evident the need to improve the accuracy of the estimation by using estimation techniques supported in machine learning that have been shown to facilitate and improve the performance of this.

Portable RDFT-Based EIS System Design With a Low-Complexity Impedance Calculation

This paper presents a novel low-complexity impedance calculation based on a compact recursive discrete Fourier transform (RDFT) for electrical impedance spectroscopy (EIS) system design. Compared to a traditional fast Fourier transform, an arbitrary-length RDFT is not limited by the power-of-two transform length. By choosing a suitable integer ratio of the desired frequency/frequency resolution, the spectrum leakage can be significantly reduced by 4-Hz resolution per bin, which enables a lower magnitude relative error (MRE) and a phase absolute error (PAE) in impedance calculations. Here, the practical measurement results show that the maximum MRE and PAE are below 0.1% and 0.5°, respectively, making this approach superior to previously reported ones. Compared with the state-of-the-art Goertzel filter-based EIS system, the proposed method provides more frequency bins (+43.75%) within the same frequency band, and the operations in terms of square root, division, and arctangent are drastically reduced by 50%, 33%, and 50%, respectively. In addition, the proposed system only costs $474 USD and can serve as a suitable solution for various EIS applications in the future.

Prediction of Software Effort by Using Non-Linear Power Regression for Heterogeneous Projects Based on Use case Points and Lines of code

Software cost estimation is necessary for all software organizations for contract negotiation. Numerous researchers have worked over the last four decades to provide estimation models, which can estimate efforts as accurately as they could, but the continuous change in the development models and use of new programming languages requires the development of new techniques for effort estimation. The size of the project must be estimated to determine the effort involved, which is rely on the programming languages and techniques employed for its development. Use case points, object points, lines of code, and story points are a few of the size matrices that are used for estimation using algorithmic models. There are several estimating techniques for the estimation of algorithmic models for various development approaches, but none of them offers a unified model for effort estimation for procedural and object-oriented projects. This work identifies common parameters for object-oriented and procedural-oriented programming languages. After the identification of common parameters, a combined model is developed for the estimation of effort for object-oriented and procedural-oriented programming language projects. The model's input is the size, which is measured in use case points and lines of source code. Power regression analysis of non-linear regression is applied to design the proposed model is designed using. The identified common parameters are multiplied with the mathematical equation achieved by power regression analysis to estimate the development effort for heterogeneous projects. The proposed work is assessed with the mean magnitude of relative error (MMRE). The error calculated with the proposed model is -0.16. The values of MMRE are further compared with the existing techniques such as linear regression, case-based reasoning, ensemble model, radial basis function neural network, and fuzzy analytic hierarchy process, and it was perceptible from the result that the proposed model delivers minimum error. The proposed work delivers the highest accuracy for software project estimation with the least amount of error.

Estimasi Pengembangan Perangkat Lunak Dengan Use Case Size Point

In software development activities one of the stages that need to be known by the project manager and the client is the estimated effort. The estimation activity is used as a reference in determining the time and personnel involved in software development. There are several methods used in software estimation such as Function Point, COCOMO, and Use Case Point. However, there are still some shortcomings such as the subjective tendency to measure complexity and lack of calculation on technical factors and environmental factors. The estimation used in this study uses the Use Case Size Point method which is the development of the Use Case Point method. The results of the Use Case size Point were used to provide a more accurate evaluation of effort with 0.01 Mean Magnitude of Relative Error (MMRE) and 0.01 Mean Magnitude of Relative Error (MMER)