Multiple Regression Algorithm Research Articles

Clinical evaluation of a first trimester algorithm predicting the risk of hypertensive disease of pregnancy

The aim of this study is to validate the Fetal Medicine Foundation (FMF) multiple logistic regression algorithm for prediction of risk of pre-eclampsia in an Australian population. This model, which predicts risk using the population rate of pre-eclampsia, a variety of demographic factors, mean maternal arterial blood pressure (MAP), uterine artery PI (UtA PI) and pregnancy-associated plasma protein A (PAPP-A), has been shown to predict early-onset pre-eclampsia (delivery prior to 34weeks) in 95% of women at a 10% false-positive rate. All women who attended first trimester screening at the Royal Prince Alfred Hospital had their body mass index (BMI), MAP and UtA PI assessed in addition to factors traditionally used to assess aneuploidy (including PAPP-A MoM). After delivery, risks of early-onset (delivery prior to 34weeks) pre-eclampsia, late pre-eclampsia and gestational hypertension were calculated using the FMF risk algorithm. A total of 3099 women were screened and delivered locally. 3066 (98.9%) women had all data to perform pre-eclampsia screening available. This included 3014 (98.3%) women with a live birth, where risks of early pre-eclampsia were calculated. Twelve women were delivered before 34weeks because of early pre-eclampsia with a prevalence of early pre-eclampsia of 1 in 256 pregnancies. Risks generated through the use of maternal history, MAP, UtA PI and PAPP-A detected 41.7 and 91.7% of early pre-eclampsia at a false-positive rate of 5 and 10%, respectively. This study shows that the FMF early pre-eclampsia algorithm is effective in an Australian population.

Noise pollution analysis in Tehran cement plant

Background : Cement industry has many process units. Basically all of these units can be considered as a source of noise. Since noise pollution is defined based on its offensive hearing effects, the importance of the noise sources depends directly on the number of workers in the unit. Materials and Methods : An experimental study has been done at Tehran cement factory to recognize and analyze the major noise pollution source machineries. In this survey, first, the mean A-weighted sound pressure level and the number of workers in each unit have been measured and the obtained data has been analyzed in Microsoft Excel. Afterwards, based on the mentioned ISO, some experiments have been done to calculate the A-weighted acoustic power level of the crusher which emits the highest level of noise. Results: The noisiest units have been recognized and some applicable suggestions have been offered to reduce the sound exposure level. Also by calculating the sound power levels and considering the design parameters which affect the sound power generated by the crusher, an empirical equation has been presented to calculate the acoustic power level as a function of horsepower, number of hammers and the weight of each hammer. Conclusion: The acoustic measurement results show that the crusher, cement mill and row mill units are the three major noise emission sources. In another experiment, based on ANSI S1.36 and ISO 3746 guidelines, the sound power level of the crusher unit has been measured indirectly. Afterwards, by using a multiple regression algorithm and minimizing the mean square error, an empirical equation has been proposed to correlate the acoustic power level as a function of the horsepower and the mass of all hammers of the crusher.

Remotely sensed retrieval of midday air temperature considering atmospheric and surface moisture conditions

Near-surface air temperature (T a) is an important variable for various scientific communities, and many previous studies have attempted to estimate T a with high spatial resolution using remotely sensed data. The present study proposes a new, practical method for estimating T a near the land surface based on moisture conditions of the atmosphere and surface. First, match-up data sets of ground observations from 76 meteorological stations and satellite remote-sensing observations were established over the Republic of Korea during 2006. Four cases of atmospheric and surface moisture conditions were then determined using Moderate Resolution Imaging Spectroradiometer brightness temperatures (at 11 and 12 μm) and the normalized difference water index for all match-ups. Using stepwise multiple regression analysis for each case, land surface temperature and the normalized difference vegetation index were statistically selected as significant independent variables for estimating T a using the development data set (75% of the match-ups for each case). The T a values estimated from the multiple regression algorithm with different constants and coefficients for each case showed reasonably good performance (nearly zero mean bias and approximately 3°C root mean square error) compared with the measured T a values used as the validation data set (25% of the match-ups for each case). Thus, the proposed T a estimation approach based on moisture conditions of the atmosphere and surface is not confined to a specific season or land type and can be applied to all areas and seasons in the Republic of Korea.

Clinical Information Research Based on Multiple Stepwise Regression Algorithm

In view of the high incidence of myopia in young people, by multiple regression analysis algorithm applied to acupuncture treatment of young patients with myopia in all the quantitative analysis of influencing factors, we found that patients get the best return for each The equation is not the same, indicating that a variety of reasons of visual acuity, and analyzes the various factors in the equation and the proportion appears in all equations, in the hope of in-depth study through a more comprehensive vision to grasp the changes caused by the substance of meaning.

Analysis to the Influences of Cutting Force to Machining Deformation of a Titanium Alloy Thin-Wall Tube

This paper took a titanium alloy thin-wall splined tube as the object of research, and mainly analyzed the influences of cutting force to machining deformation of part in turning process. The paper first introduced the structure composition and characteristic of splined tube, then established the mathematical model of cutting force in turning process according to the orthogonal experiment method and the multiple regression algorithm, and finally the influences of cutting force to machining deformation of splined tube was carried on the simulation analysis with the aid of the finite element analysis software under the different cutting conditions. The conclusions of research in this paper will have important practical significance to reduce the machining deformation of thin-wall part and optimize the cutting parameters of difficult-to-machine material.

A stochastic model accommodating the FAMACHA © system for estimating worm burdens and associated risk factors in sheep naturally infected with Haemonchus contortus

A previously developed multiple regression algorithm was used as the basis of a stochastic model to simulate worm burdens in sheep naturally infected with Haemonchus contortus over five consecutive Haemonchus seasons (November to January/February) on a farm in the summer rainfall region in South Africa, although only one season is discussed. The algorithm associates haemoglobin levels with worm counts in individual animals. Variables were represented by distributions based on FAMACHA © scores and body weights of sheep, and Monte Carlo sampling was used to simulate worm burdens. Under conditions of high disease risk, defined as the sampling event during the worm season with the lowest relative mean haemoglobin level for a class of sheep, the model provided a distribution function for mean class H. contortus burdens and the probability of these occurring. A mean H. contortus burden for ewes ( n = 130 per sample) of approximately 1000 (range 51–28,768) and 2933 (range 78–44,175) for rams ( n = 120 per sample) was predicted under these conditions. At the beginning of the worm season when the risk of disease was lowest (i.e. when both classes had their highest estimated mean haemoglobin levels), a mean worm burden of 525 (range 39–4910) for ewes and 651 (range 37–17,260) for rams was predicted. Model indications were that despite being selectively drenched according to FAMACHA © evaluation, 72% of the ewes would maintain their mean worm burden below an arbitrarily selected threshold of 1000 even when risk of disease was at its highest. In contrast, far fewer rams (27%) remained below this threshold, especially towards the end of the worm season. The model was most sensitive to changes in haemoglobin value, and thus by extrapolation, the haematocrit, which is used as the gold standard for validating the FAMACHA © system. The mean class haemoglobin level at which there was a 50% probability of worm burdens being ≤1000 worms was 7.05 g/dl in ewes and 7.92 g/dl in rams.

MICR Multiple Instance Cluster Regression for Student Academic Performance in Higher Education

is an increasing need for the analysis and prediction of the student academic performance in higher education. The ability to predict the student academic performance is also most important in higher education system. There are many challenges in this regard. In this paper, we present a Multiple Instance regression algorithm that models the internal data structure of three data sets to identify items that are most relevant to the data set labels, which operates on three data sets with real valued labels, each containing a set of unlabeled items, in which the relevance of each item to its data set is unknown. The goal is to predict the labels of new data set from its contents. Unlike previous Multiple instance regression methods, Multiple instance cluster Regression can operate on datasets that are structured in which they contain items drawn from a number of distinct unknown distributions. Multiple instance cluster Regression provided predictions that were more accurate than those obtained with non multiple instance approaches or multiple instance regression methods that do not model the data set structures. This Paper makes an attempt towards the application of multiple instance regression algorithms to decide to which category a student belongs (i.e, low#risk students, medium#risk students, high#risk students).

A Parallel Implementation of the Network Identification by Multiple Regression (NIR) Algorithm to Reverse-Engineer Regulatory Gene Networks

The reverse engineering of gene regulatory networks using gene expression profile data has become crucial to gain novel biological knowledge. Large amounts of data that need to be analyzed are currently being produced due to advances in microarray technologies. Using current reverse engineering algorithms to analyze large data sets can be very computational-intensive. These emerging computational requirements can be met using parallel computing techniques. It has been shown that the Network Identification by multiple Regression (NIR) algorithm performs better than the other ready-to-use reverse engineering software. However it cannot be used with large networks with thousands of nodes - as is the case in biological networks - due to the high time and space complexity. In this work we overcome this limitation by designing and developing a parallel version of the NIR algorithm. The new implementation of the algorithm reaches a very good accuracy even for large gene networks, improving our understanding of the gene regulatory networks that is crucial for a wide range of biomedical applications.

Prediction of joint center location by customizable multiple regressions: Application to clavicle, scapula and humerus

Accurate spatial location of joint center (JC) is a key issue in motion analysis since JC locations are used to define standardized anatomical frames, in which results are represented. Accurate and reproducible JC location is important for data comparison and data exchange. This paper presents a method for JC locations based on the multiple regression algorithms without preliminary assumption on the behavior of the joint-of-interest. Regression equations were obtained from manually palpable ALs on each bone-of-interest. Results are presented for all joint surfaces found on the clavicle, scapula and humeral bone. Mean accuracy errors on the JC locations obtained on dry bones were 5.2±2.5 mm for the humeral head, 2.5±1.1 mm for the humeral trochlea, 2.3±0.9 mm for the humeral capitulum, 8.2±3.9 mm for the scapula glenoid cavity, 7.2±3.2 mm for the scapular aspect of the acromio-clavicular joint, 3.5±1.8 mm for the clavicular aspect of the sternoclavicular joint and 3.2±1.4 mm for the clavicular aspect of the acromio-clavicular joint. In-vitro and in-vivo validation accuracy was 5.3 and 8.5 mm, respectively, for the humeral head center location. Regression coefficients for joint radius dimension and joint surface orientation were also processed and reported in this paper.

Water quality monitoring in a slightly-polluted inland water body through remote sensing — Case study of the Guanting Reservoir in Beijing, China

This study focused on the water quality of the Guanting Reservoir, a possible auxiliary drinking water source for Beijing. Through a remote sensing (RS) approach and using Landsat 5 Thematic Mapper (TM) data, water quality retrieval models were established and analyzed for eight common water quality variables, including algae content, turbidity, and concentrations of chemical oxygen demand, total nitrogen, ammonia nitrogen, nitrate nitrogen, total phosphorus, and dissolved phosphorus. The results show that there exists a statistically significant correlation between each water quality variable and remote sensing data in a slightly-polluted inland water body with fairly weak spectral radiation. With an appropriate method of sampling pixel digital numbers and multiple regression algorithms, retrieval of the algae content, turbidity, and nitrate nitrogen concentration was achieved within 10% mean relative error, concentrations of total nitrogen and dissolved phosphorus within 20%, and concentrations of ammonia nitrogen and total phosphorus within 30%. On the other hand, no effective retrieval method for chemical oxygen demand was found. These accuracies were acceptable for the practical application of routine monitoring and early warning on water quality safety with the support of precise traditional monitoring. The results show that performing the most traditional routine monitoring of water quality by RS in relatively clean inland water bodies is possible and effective.

Multiple regression with fuzzy data

In this paper, we propose an iterative algorithm for multiple regression with fuzzy variables. While using the standard least-squares criterion as a performance index, we pose the regression problem as a gradient-descent optimisation. The separation of the evaluation of the gradient and the update of the regression variables makes it possible to avoid undue complication of analytical formulae for multiple regression with fuzzy data. The origins of fuzzy input data are traced back to the fundamental concept of information granulation and an example FCM-based granulation method is proposed and illustrated by some numerical examples. The proposed multiple regression algorithm is applied to one-, three- and nine-dimensional synthetic data sets as well as the 13-dimensional Boston Housing dataset from the machine learning repository. The algorithm's performance is illustrated by the corresponding plots of convergence of regression parameters and the values of the prediction error of the resulting regression model. General comments on the numerical complexity of the proposed algorithm are also provided.

Water Quality Retrievals from High Resolution Ikonos Multispectral Imagery: A Case Study in Istanbul, Turkey

This paper presents an application of high resolution satellite remote sensing data for mapping water quality in the Goldon Horn, Istanbul. It is an applied research emphasizing the present water quality conditions in this region for water quality parameters; secchi disc depth (SDD), chlorophyl-a (chl-a) and total suspended sediment (TSS) concentration. The study also examines the retrievals of these parameters through high resolution IKONOS multispectral data supported by in situ measurements. Image processing procedure involving radiometric correction is carried out for conversion from digital numbers (DNs) to spectral radiance to correlate water quality parameters and satellite data by using multiple regression technique. The retrieved and verified results show that the measured and estimated values of water quality parameters in good agreement (R 2 > 0.97). The spatial distribution maps are developed by using multiple regression algorithm belonging to water quality parameters. These maps present apparent spatial variations of selected parameters and inform the decision makers of water quality variations in a large water region in the Istanbul metropolitan area.

Contribution of multispectral satellite imagery to the bathymetric analysis of coastal sea bottom

The knowledge of the topography of the seafloor is important for several applications. Image satellites of observation of the earth are the result of the interaction of the electromagnetic radiance with the system sea - atmosphere and sensor. To understand this complex phenomena us developed an analytic model of radiatif transfer simulation in water coupled to an atmospheric model in order to simulate measure by satellite. This direct model permits to follow the solar radiance in his trajectory Sun-Atmosphere - Sea - Depth of sea- sensor. The goal of this simulation is to show for every satellite of observation (Alsat1, Spot, Landsat MSS, Landsat TM) possibilities that can offer in domain of bathymetry. The principle of measure of bathymetry necessarily takes this model of reflectance joining the intensity of radiometric signal measured by the satellite to the depth as a basis; it can call on the physical method that requires the knowledge of all parameters governing this model (optic properties of water, coefficient of reflection of the bottom, transmittance of the atmosphere,). In application, a set of relations figures are gotten to determine the bathymetry of an inshore zone of Algeria. The model provides of image mono channel where each pixel of the maritime domain is represented either by a radiometry in - situ but rather by a calculated depth. In general the use of hybrid multiple Spot band regression algorithms are superior to the exclusive use of any single band.

Nondestructive materials characterization of irradiated nuclear pressure-vessel steel samples by the use of micromagnetic techniques and in terms of mechanical properties

The contribution reports to the state of the art in materials characterization obtained by the authors through the use of micromagnetic nondestructive testing (NDT) techniques in order to predict mechanical properties of steel samples (surveillance program) after neutron irradiation in irradiation chambers of nuclear pressure vessels. The investigations have been performed within the framework of the EURATOM project CRETE [1]. The irradiated steel samples were delivered by different European countries: two sets of specimen came from France, two came from Germany, two from the Czech Republic and one from Spain. The specimens were stored in the hot cell of the research reactor in Petten, Netherlands, where they could be handled by robots in order to perform the measurements by different inspection teams. Besides electromagnetic (one team) and micromagnetic (two teams) techniques, thermoelectric (2 teams) techniques have been applied and the results are discussed. In order to characterize microstructure states in ferromagnetic steels micromagnetic, NDT techniques were based on the interaction of Bloch walls with lattice defects such as vacancies, dissolved atoms, precipitates, dislocations, and grain-and phase-boundaries. The way the lattice defects are the pinning points of the dislocation movement under mechanical loads, i.e., how they contribute to a strengthening effect, is similar to the way the Bloch walls interact with the lattice defects under magnetic loads. In addition, magnetic strengthening is observed that can be detected, for instance, in an increase of the coercivity. The institution of the authors has developed the so-called 3MA-approach (micromagnetic, multiple-parameter, microstructure, and stress analysis) in the last decade. This approach combines the information resulting from the performance of different micromagnetic techniques (magnetic Barkhausen noise, incremental permeability, harmonic analysis of the magnetic tangential field and eddy current testing used at 3 different frequencies). The data fusion is obtained through the use of neural networks, multiple regression algorithms, or pattern recognition procedures such as the nearest neighbor approach. Various information in a calibration process is selected such that disturbing influences, for instance, heating and cooling effects, are suppressed and the signal-to-noise ratio for the target function prediction (hardness, yield strength, Charpy energy, fracture appearance transition temperature, etc.) is optimized. 3MA is performed by an intelligent transducer and a data evaluation system with which, in a time multiplexing mode, one measuring quantity after the other is selected. The results obtained at the broken halves of ISO-V-notched Charpy specimens show a good performance of the multiple regression approach, with high regression coefficients near 1 and a small residual standard deviation between modeled and destructively determined mechanical properties. However, stochastically independently selected specimens show a larger deviation than the standard error of the calibration. The reason for this phenomenon is that the plastic deformation and the residual stress are in the broken Charpy specimen. Therefore, further investigations will be concentrated on calibration on irradiated, but unbroken, Charpy specimens.

The development of an in-process surface roughness adaptive control system in end milling operations

An in-process surface roughness adaptive control (ISRAC) system in end milling operations was researched and developed. A multiple regression algorithm was employed to establish two subsystems: the in-process surface roughness evaluation (ISRE) subsystem and the in-process adaptive parameter control (IAPC) subsystem. These systems included not only machine cutting parameters such as feed rate, spindle speed, and depth of cut, but also cutting force signals detected by a dynamometer sensor. The multiple-regression-based ISRE subsystem predicted surface roughness during the finish cutting process with an accuracy of 91.5%. The integration of the two subsystems led to the ISRAC system. The testing resulted in a 100% success rate for adaptive control, proving that this proposed system could be implemented to adaptively control surface roughness during milling operations. This research suggests that multiple linear regression used in this study was straightforward and effective for in-process adaptive control.

Multiple Testing and Data Adaptive Regression: An Application to HIV-1 Sequence Data.

Analysis of viral strand sequence data and viral replication capacity could potentially lead to biological insights regarding the replication ability of HIV-1. Determining specific target codons on the viral strand will facilitate the manufacturing of target-specific antiretrovirals. Various algorithmic and analysis techniques can be applied to this application. In this paper, we apply two techniques to a data set consisting of 317 patients, each with 282 sequenced protease and reverse transcriptase codons. The first application is recently developed multiple testing procedures to find codons which have significant univariate associations with the replication capacity of the virus. A single-step multiple testing procedure (Pollard and van der Laan 2003) method was used to control the family wise error rate (FWER) at the five percent alpha level as well as the application of augmentation multiple testing procedures to control the generalized family wise error (gFWER) or the tail probability of the proportion of false positives (TPPFP). We also applied a data adaptive multiple regression algorithm to obtain a prediction of viral replication capacity based on an entire mutant/non-mutant sequence profile. This is a loss-based, cross-validated Deletion/Substitution/Addition regression algorithm (Sinisi and van der Laan 2004), which builds candidate estimators in the prediction of a univariate outcome by minimizing an empirical risk. These methods are two separate techniques with distinct goals used to analyze this structure of viral data.

Effects of pore aspect ratios on velocity prediction from well‐log data

ABSTRACTWe develop a semi‐empirical model which combines the theoretical model of Xu and White and the empirical formula of Han, Nur and Morgan in sand–clay environments. This new model may be used for petrophysical interpretation of P‐ and S‐wave velocities. In particular, we are able to obtain an independent estimation of aspect ratios based on log data and seismic velocity, and also the relationship between velocities and other reservoir parameters (e.g. porosity and clay content), thus providing a prediction of shear‐wave velocity. To achieve this, we first use Kuster and Toksöz's theory to derive bulk and shear moduli in a sand–clay mixture. Secondly, Xu and White's model is combined with an artificial neural network to invert the depth‐dependent variation of pore aspect ratios. Finally these aspect ratio results are linked to the empirical formula of Han, Nur and Morgan, using a multiple regression algorithm for petrophysical interpretation. Tests on field data from a North Sea reservoir show that this semi‐empirical model provides simple but satisfactory results for the prediction of shear‐wave velocities and the estimation of reservoir parameters.

Path Corrections for Source Discriminants: A Case Study at Two International Seismic Monitoring Stations

Improving the performance of short-period regional seismic discriminants by applying propagation corrections is explored using observations from two seismic monitoring stations in Asia. Frequency-dependent regional phase amplitude ratio measurements at stations NIL and ZAL for earthquakes and underground nuclear explosions were obtained from the prototype-International Data Center (pIDC) that has been established for developing monitoring capabilities of the Comprehensive Nuclear-Test-Ban Treaty (CTBT). The pIDC discriminant measurements have large scatter, much of which is attributed to wave propagation effects in the heterogeneous crustal waveguide. Linear regressions indicate that the phase ratios are correlated with topographic characteristics along the individual paths, providing an empirical means for correcting for path effects beyond conventional distance corrections. Kriging, a spatial multiple regression algorithm, also reveals coherent spatial patterns in the data indicative of regional path effects. Using available high-resolution topography data, correction of regional P/S ratios for the best models obtained from multivariate regressions systematically reduces the data variance relative to distance corrections alone, as has been observed for other data sets. The reduced scatter in the measurements increases the separation between earthquake and explosion populations in most cases, enhancing the regional discriminant performance. The path-corrected discriminants isolate explosions better for NIL than for ZAL, even though some of the explosion sources are located in a common source area. Kriging achieves comparable or superior variance reduction for the discriminant measures, without requiring knowledge of the path structure, although this may not result in improved discriminant performance. While always desirable, corrections for heterogeneous path effects may prove inadequate in some cases, notably when phase blockage occurs or when strong attenuation eliminates the diagnostic high-frequency energy.

Potential of combined spaceborne infrared and microwave radiometry for near real‐time rainfall attenuation monitoring along earth‐satellite links

AbstractThe objective of this paper is to investigate how spaceborne remote sensors, and their derived products, can be exploited to optimize the performances of a satellite communication system in the presence of precipitating clouds along the path. The complementarity between sun‐synchronous microwave (MW) and geo‐stationary infrared (IR) radiometry for monitoring the earth's atmosphere is discussed and their potential as a rain detection system within near real‐time countermeasure techniques for earth‐satellite microwave links is analysed. A general approach, consisting in estimating rainfall intensity and attenuation by polar‐orbiting microwave radiometers and temporally tracking the rainfall areas by geo‐stationary infrared radiometers, is delineated. Multiple regression algorithms for predicting rainfall attenuation from spaceborne brightness temperatures and from surface rainrate, trained by radiative transfer and cloud models, are illustrated. A maximum likelihood technique is delineated to discriminate stratiform and convective rainfall from spaceborne brightness temperatures. The differences among attenuation estimates derived from layered raining‐cloud structures with respect to those obtained from simple rain slabs, as recommended by ITU‐R, are also quantified. A test of the proposed attenuation prediction methods is performed using raingage and Italsat data acquired in Spino d'Adda (Italy) during 1994. A description of the statistical method, based on the probability matching technique, adopted to combine MW and IR data for retrieving and tracking precipitating cloud systems in terms of path attenuation and accumulated rain at ground is finally provided together with its application to a case study over the Mediterranean area during October 1998. Copyright © 2001 John Wiley & Sons, Ltd.

Specific versus Nonspecific Brain Activity in a Parametric N-Back Task

In this study functional magnetic resonance imaging (fMRI) was used to examine cerebral activity patterns in relation to increasing mental load of a working memory task. Aim of the experiment was to distinguish nonspecific task-related processes from specific workload processes analytically. Twelve healthy volunteers engaged in a spatial n-back task with four levels. FMRI data were acquired with the 3D-PRESTO pulse sequence. Analysis entailed a two-step multiple regression algorithm, which was specifically designed to measure and separate load-sensitive and load-insensitive activity simultaneously, while preserving the original high spatial resolution of the fMRI signal. Load-sensitive and load-insensitive activity was found in both dorsolateral-prefrontal and parietal cortex, predominantly bilaterally, and in the anterior cingulate. As expected, the left primary sensorimotor cortex showed predominantly load-insensitive activity. Load-sensitive activity reflects specific working memory functions, such as temporary retention and manipulation of information, while load-insensitive activity reflects supportive functions, such as visual orientation, perception, encoding, and response selection and execution. Good performance was correlated with a large area of load-sensitive activity in anterior cingulate, and with a small area of load-insensitive activity in the right parietal cortex. The findings indicate that nonspecific and specific working memory processes colocalize and are represented in multiple frontal and parietal regions. Implication of this analytical strategy for application in research on psychiatric disorders is discussed.