Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “Linear Regression”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 595 records · Page 33Linked to original sources
PubMed1992

Gene-environment interaction in hypertension, stroke and atherosclerosis in experimental models and supportive findings from a world-wide cross-sectional epidemiological survey: a WHO-cardiac study.

1. Genetic rat models of hypertension, such as spontaneously hypertensive rats (SHR) and in particular, stroke-prone SHR (SHRSP) are useful models for research on the genetic pathogenesis, gene-environment interaction and control of environmental factors for the prevention of hypertension and related cardiovascular diseases (CVD). 2. Since recent genetic analysis of hypertension in SHRSP indicated that one of the hypertensive genes related to salt-induced blood pressure (BP) rise was linked with the gene of the angiotensin-converting enzyme, gene-environment interaction is important in the humoral, neural, vascular and nutritional mechanisms of hypertension and CVD. 3. Extensive experimental studies in SHRSP by the authors have demonstrated nutritional factors counteracting directly or indirectly against the adverse effect of excess salt intake, such as K, Mg, Ca, dietary fibres, protein, some amino acids and fatty acids, etc.; they are therefore effective in preventing stroke, the typical complication of hypertension. 4. Experimental atherosclerotic models were established in SHRSP, which indicated also a gene-environment interaction in the pathogenesis of atherosclerosis at the cellular level of vascular smooth muscle cells. Excess salt intake accelerated cholesterol absorption from the intestine to induce arterial fat deposition as well as to active platelet aggregation by the mechanism of increased intracellular Ca2+ mobilization. 5. Based on such experimental findings on the hypertension and related CVD prevention, a cross-sectional multicentre epidemiological 'CVD and Alimentary Comparison' Study (WHO-CARDIAC Study) was designed to assess the relationship of biological markers of dietary factors with BP ('core' study) and major CVD mortalities ('complete' study) and has been successfully undertaken for the past 8 years in co-operation with 54 centres in 23 countries. 6. The results of 'core' study so far obtained by cross-centre simple linear regression analysis demonstrated adverse effects of body mass index and salt intake on BP in men and possible beneficial effects of Mg by within-centre multiple linear regression analysis. 7. Preliminary cross-centre simple linear regression analysis of the 'complete' study indicated a positive correlation of serum cholesterol levels and negative correlation of serum phospholipid poly-unsaturated to unsaturated fatty acid ratios and 24 h urinary taurine excretion with coronary heart diseases mortality; a positive correlation of 24 h urinary Na and Na/K ratios and a negative correlation of serum cholesterol levels with stroke mortalities was indicated. 8. These epidemiological data confirmed mostly the applicability of experimental findings to the nutritional prevention of hypertension and CVD prevention.(ABSTRACT TRUNCATED AT 400 WORDS)

Animals↗
PubMed2004

Adaptive ventilator FiO2 advisor: use of non-invasive estimations of shunt.

A non-invasive and simple method of parameter estimation has been developed for the model-based decision support of the artificial ventilation in intensive care units. The parameter concerned was the respiratory shunt. Originally, the shunt had to be estimated using a numerical algorithm, which was slow and unreliable. The estimation process also required the knowledge of other parameters, whose values could only be obtained using invasive monitoring equipment. In this paper, the respiratory index is used for the shunt estimation. A linear regression model and a non-linear adaptive neuro-fuzzy inference system (ANFIS) model were used to describe the relationship between the respiratory index and the shunt. The shunts estimated using these models were then used to calculate the fractional inspired oxygen needed to attain the target arterial oxygen level of the model patient. The advisor also utilises population median values of the cardiac index and oxygen consumption index. This alleviates the need for invasive monitoring. In a simulation study, the mean squared error of the control using the ANFIS model was 0.75 kPa2 compared to 2.06 kPa2 using the linear regression model. Therefore, the performance of the FiO2 advisor was better when the shunt was estimated using the non-linear ANFIS model.

Algorithms↗
PubMed1999

Feature extraction of event-related potentials using wavelets: an application to human performance monitoring.

This report describes the development and evaluation of mathematical models for predicting human performance from discrete wavelet transforms (DWT) of event-related potentials (ERP) elicited by task-relevant stimuli. The DWT was compared to principal components analysis (PCA) for representation of ERPs in linear regression and neural network models developed to predict a composite measure of human signal detection performance. Linear regression models based on coefficients of the decimated DWT predicted signal detection performance with half as many free parameters as comparable models based on PCA scores. In addition, the DWT-based models were more resistant to model degradation due to over-fitting than PCA-based models. Feed-forward neural networks were trained using the backpropagation algorithm to predict signal detection performance based on raw ERPs, PCA scores, or high-power coefficients of the DWT. Neural networks based on high-power DWT coefficients trained with fewer iterations, generalized to new data better, and were more resistant to overfitting than networks based on raw ERPs. Networks based on PCA scores did not generalize to new data as well as either the DWT network or the raw ERP network. The results show that wavelet expansions represent the ERP efficiently and extract behaviorally important features for use in linear regression or neural network models of human performance. The efficiency of the DWT is discussed in terms of its decorrelation and energy compaction properties. In addition, the DWT models provided evidence that a pattern of low-frequency activity (1 to 3.5 Hz) occurring at specific times and scalp locations is a reliable correlate of human signal detection performance.

Evoked Potentials↗
PubMed2005

[Effects of post diameter on retention of post-core crown system].

OBJECTIVE: To access the effects of post's diameter on the retention of post-core restorations. METHODS: Thirty-three simulant roots with same length,taper and different diameter (1.305 mm, 1.830 mm, 2.175 mm and 2.700 mm respectively) were made of polymethylmethacrylate and were divided into 4 groups. The wax patterns of post-cores were manufactured and cast. The post-cores were cemented. After that, the adhesion, frictional force and restriction of the post-core restorations were measured. RESULTS: When the proportion of length and diameter was smaller than 4.372, the linear regression between the post's diameter and the retention of post-core crown system was F = 107.98 + 72.08D. The linear regression between the post's diameter and the adhesion was P = 9.60 + 2.22D. The linear regression between the post's diameter and the frictional force and restriction was Q = 98.37 + 69.87D. The statistical analysis of the results indicated significant difference (P < 0.05). When the proportion of length and diameter was lager than 4.372, the retention of post-core crown system increased although the post's diameter decreased. CONCLUSION: When the proportion of length and diameter was smaller than 4.372, the retention of post-core crown system increased with the decrease of the post's diameter. While the proportion was lager than a certain value, this rule did not exist.

Crowns↗
PubMed1993

Design effects for binary regression models fitted to dependent data.

Dependent data, such as arise with cluster sampling, typically yield variances of parameter estimates which are larger than would be provided by a simple random sample of the same size. This variance inflation factor is called the design effect of the estimator. Design effects have been derived for cluster sampling designs using simple estimators such as means and proportions, and also for linear regression coefficient estimators. In this paper, we show that a method to derive design effects for linear regression estimators extends to generalized linear models for binary responses. In particular, some simple expressions for design effects in the linear regression model provide accurate approximations for binary regression models such as those based on the logistic, probit and complementary log-log links. We corroborate our findings with two examples and some simulation studies.

Acquired Immunodeficiency Syndrome↗
PubMed1995

A sensitometric comparison of Fuji Super HR-G and Kodak T-mat G panoramic films.

Film contrast, relative speed, and resolution of Fuji Super HR-G panoramic film were determined according to specifications of the American National Standards Institute and compared with those of Kodak T-mat G panoramic film. The films were simultaneously exposed by Kodak Lanex regular intensifying screens with time-scale sensitometry. Optical density was read with a digital densitometer, and characteristic curves were plotted and analyzed by polynomial and linear regression. The average gradients of the linear regression curves between 0.25 and 2.00 optical density units above base plus fog were taken as contrast (HR-G: 1.831, T-mat G: 1.712). Relative speed, when exposed with Lanex regular intensifying screens and calculated at 1.00 optical density unit above base plus fog, was 412 for HR-G. A speed of 400 was chosen for T-mat G to match Kodak consumer information. Resolution for both films was similar. These findings suggest that Fuji Super HR-G panoramic film develops similar contrast, speed, and resolution to Kodak T-mat G panoramic film.

Linear Models↗
PubMed1998

Regression-based reference limits and their reliability: example on hemoglobin during the first year of life.

Calculation of reference limits by regression analysis makes it unnecessary to partition the reference data into subgroups, and age-dependent limits can be estimated within as narrow age intervals as necessary. However, the reliability of the regression-based reference limits has not been considered before. To get valid regression-based confidence intervals (CIs) for reference limits, one must evaluate the convolution of two distributions. In this study, age-dependent reference limits with corresponding CIs were produced for blood hemoglobin concentrations over the age interval from newborns to 12 months. We describe how the variance associated with the reference limits can be estimated, and present a Table from which appropriate values can be chosen for the calculation of regression-based reference limits and exact CIs. Also, an equation for the calculation of approximate CIs is given. The data were modeled by linear regression in several cumulative age groups to find the transition zone where the slope changed. After defining this cutoff point, piecewise linear regression was applied. Reference limits and their CIs calculated by conventional and piecewise linear regression methods were almost the same in older age groups but differed significantly during the period of most rapid age-dependent changes, i.e., during the 2 months after birth.

Age Factors↗
PubMed1991

Comparison of fractional excretion and 24-hour urinary excretion of sodium and potassium in clinically normal cats and cats with induced chronic renal failure.

The influence of induced chronic renal failure on 24-hour urinary excretion and fractional excretion of sodium and potassium was studied in cats. Induction of chronic renal failure significantly increased fractional excretion of potassium (P less than 0.0001) and sodium (P less than 0.05); however, 24-hour urinary excretion of sodium and potassium decreased slightly following induction of chronic renal failure. Fractional excretion and 24-hour urinary excretion of sodium and potassium were compared by linear regression in clinically normal cats, cats with chronic renal failure, and clinically normal and affected cats combined. In clinically normal cats, linear regression revealed only moderate correlation between fractional excretion and 24-hour urinary excretion for sodium and potassium. Linear regression of these same relationships in cats with chronic renal failure, and in clinically normal cats and cats with chronic renal failure combined, indicated low correlation. Fractional excretions of sodium and potassium were not reliable indicators of 24-hour urinary excretion of these electrolytes in cats with chronic renal failure or unknown glomerular filtration rate. Fractional excretion of potassium and sodium correlated only moderately with 24-hour urinary excretion in clinically normal cats.

Animals↗
PubMed2006

Asymmetry of right versus left intraocular pressures over 24 hours in glaucoma patients.

PURPOSE: To assess the asymmetry of 24-hour intraocular pressures (IOPs) between right and left eyes of untreated open-angle glaucoma (OAG) patients, and its influence on the monocular therapeutic trial. DESIGN: Experimental study with human patients. PARTICIPANTS: Forty-one subjects (40-78 years old) with untreated OAG. METHODS: Subjects were housed in a sleep laboratory for 24 hours. Intraocular pressures of both eyes were measured with a pneumatonometer every 2 hours with the patient in the sitting and supine positions from 7 am to 11 pm and in the supine position only from 11 pm to 7 am. Mean, peak, and trough IOPs were compared in right versus left eyes. The strength of association between right and left IOPs was assessed using coefficients of determination (R2). Validity of the monocular therapeutic trial was assessed by examining residual values from 2 linear regression models of right versus left IOP. MAIN OUTCOME MEASURES: Strength of association between IOPs of right and left eyes, and residual values from linear models of IOP. RESULTS: No statistically significant difference was found between mean, peak, or trough IOPs of right and left eyes over a 24-hour period. The strength of association for mean IOP was only moderate (R(2) = 0.421-0.623). Residual values of > or =3 mmHg were found in 14.0%+/-12.0% (mean +/- standard deviation) of IOP measurements for a symmetric linear regression model, and 8.5%+/-10.6% of IOP measurements for a best-fit linear regression model over 24 hours. CONCLUSIONS: In this group of untreated glaucoma patients, the strength of association between the right and left eye mean IOPs is only moderate. Residual values of > or =3 mmHg were more common using a symmetric model than with a best-fit model. The current method of performing monocular therapeutic trials in glaucoma patients is unreliable, but it may be possible to improve reliability by using several IOP measurements obtained at different times of the day instead of a single office measurement.

Adult↗
PubMed2004

Estimation of the density of nymphs of the bush tick, Haemaphysalis longicornis (Acari: Ixodidae), by the catch effort method.

The density of nymphs of the bush tick, Haemaphysalis longicornis, was investigated by the catch effort method with flagging. The spatial distribution of H. longicornis nymphs fit the model of contagiously distributed colonies by Iwao's m*-m analysis (Iwao 1968). A sequential sampling method was used to predict the theoretical point at which to stop sampling. Our analysis showed that five quadrats (4 m x 4 m) were sufficient to estimate the density of H. longicornis nymphs with a mean density of 5.39 per quadrat. We estimated the tick density by two methods with respect to the sampling interval. The estimated density of ticks based on ticks collected during short sampling intervals (within a half hour) was 511.34 in the 18 quadrats with the extrapolation of the linear regression equation. On the other hand, for the long interval sampling, the total number of ticks estimated by the linear regression equation was 635.47 in six quadrats in which ticks had been collected by long interval sampling. There was a significant difference between the slopes of the two linear regression equations, suggesting that the rate of reduction in the number of H. longicornis nymphs in the study area by the catch effort method differed between the two sampling methods.

Animals↗
PubMedDate not supplied

Statistical modeling of a ligand knowledge base.

A range of different statistical models has been fitted to experimental data for the Tolman electronic parameter (TEP) based on a large set of calculated descriptors in a prototype ligand knowledge base (LKB) of phosphorus(III) donor ligands. The models have been fitted by ordinary least squares using subsets of descriptors, principal component regression, and partial least squares which use variables derived from the complete set of descriptors, least angle regression, and the least absolute shrinkage and selection operator. None of these methods is robust against outliers, so we also applied a robust estimation procedure to the linear regression model. Criteria for model evaluation and comparison have been discussed, highlighting the importance of resampling methods for assessing the robustness of models and the scope for making predictions in chemically intuitive models. For the ligands covered by this LKB, ordinary least squares models of descriptor subsets provide a good representation of the data, while partial least squares, principal component regression, and least angle regression models are less suitable for our dual aims of prediction and interpretation. A linear regression model with robustly fitted parameters achieves the best model performance over all classes of models fitted to TEP data, and the weightings assigned to ligands during the robust estimation procedure are chemically intuitive. The increased model complexity when compared to the ordinary least squares linear model is justified by the reduced influence of individual ligands on the model parameters and predictions of new ligands. Robust linear regression models therefore represent the best compromise for achieving statistical robustness in simple, chemically meaningful models.

Algorithms↗
PubMed2000

The importance of subterranean mosquito habitat to arbovirus vector control strategies in north Queensland, Australia.

In north Queensland, 14 localities were surveyed for mosquito larvae (third and fourth instar) during summer/autumn and winter from 1996 to 1999. Absolute population numbers in subterranean habitats, mainly service manholes and pits (97%) but also some wells, septic tanks, storm drains, and sumps, were expressed as a proportion of total numbers in these sites plus surface sites within a 100-m radius. When correction factors were applied to subterranean samples, the 472,477 larvae mainly of Aedes tremulus (Theobald) group, Aedes notoscriptus (Skuse), and Aedes aegypti (L.) comprised 78% of the total population. In relation to the proportion of the overall immature mosquito population from subterranean habitats (propsub), linear regression coefficients for minimum temperature, relative humidity, and Mesocyclops copepod prevalence were significant for winter data; but for summer, only relative humidity was significant. Linear regression coefficients for Mesocyclops prevalence approached significance (P = 0.061) in summer. When multiple linear regression was used to model propsub, 68% of the variation was accounted for by relative humidity and the prevalence of Mesocyclops. In the drier and cooler towns, increased use of subterranean sites during winter was caused by reduced availability of surface oviposition sites because of the dry season. In the wetter coastal towns, no such restrictions applied and ambient conditions remained more equitable all year round. Mesocyclops were surprisingly common, particularly in these coastal towns. Release of known numbers of Mesocyclops indicated that 3-sweep netting in service manholes was sensitive down to densities of one Mesocyclops per 10 liters, and overall recovery varied from 1 to 4%. In relation to control, service manholes represent a stable habitat for mosquito (7% positive overall) and Mesocyclops populations. If they remained wet, service manholes positive for mosquito immatures or Mesocyclops during summer/autumn had 96% and 85% chance, respectively, of being positive the following winter. Even allowing for the effect of drying, a mosquito-positive manhole had a 79% chance of remaining positive the following winter. In view of the importance of these sites as refuges from adverse ambient conditions, it is proposed that a winter control strategy using Mesocyclops presents a cost-effective control option to reduce the recolonization of surface sites when conditions become more suitable.

Aedes↗
PubMed2004

The use of bootstrap methods for analysing Health-Related Quality of Life outcomes (particularly the SF-36).

Health-Related Quality of Life (HRQoL) measures are becoming increasingly used in clinical trials as primary outcome measures. Investigators are now asking statisticians for advice on how to analyse studies that have used HRQoL outcomes.HRQoL outcomes, like the SF-36, are usually measured on an ordinal scale. However, most investigators assume that there exists an underlying continuous latent variable that measures HRQoL, and that the actual measured outcomes (the ordered categories), reflect contiguous intervals along this continuum. The ordinal scaling of HRQoL measures means they tend to generate data that have discrete, bounded and skewed distributions. Thus, standard methods of analysis such as the t-test and linear regression that assume Normality and constant variance may not be appropriate. For this reason, conventional statistical advice would suggest that non-parametric methods be used to analyse HRQoL data. The bootstrap is one such computer intensive non-parametric method for analysing data. We used the bootstrap for hypothesis testing and the estimation of standard errors and confidence intervals for parameters, in four datasets (which illustrate the different aspects of study design). We then compared and contrasted the bootstrap with standard methods of analysing HRQoL outcomes. The standard methods included t-tests, linear regression, summary measures and General Linear Models.Overall, in the datasets we studied, using the SF-36 outcome, bootstrap methods produce results similar to conventional statistical methods. This is likely because the t-test and linear regression are robust to the violations of assumptions that HRQoL data are likely to cause (i.e. non-Normality). While particular to our datasets, these findings are likely to generalise to other HRQoL outcomes, which have discrete, bounded and skewed distributions. Future research with other HRQoL outcome measures, interventions and populations, is required to confirm this conclusion.

Arthritis, Rheumatoid↗
PubMed1990

Characteristics of pulmonary diffusing capacity for CO at rest and during exercise.

Characteristic patterns of changes in pulmonary diffusing capacity (DL) at rest and during exercise were investigated and characteristics of normal DL values concerned on sex, age, and ethnic groups were examined by viewing our studies and other reports. The relation of DL and pulmonary capillary blood volume (Vc) was represented as a logarithmic regression at rest and as a linear regression during exercise. The curve relation at rest is considered to show that the increase in Vc mainly reflects the process of transport from pulmonary capillary recruitment to pulmonary capillary dilation. The increasing rate of DL was not decreased during exercise, which seemed to be due to an increase in pulmonary blood flow accompanying exercise. The linear regression was also found between DL and oxygen intake during exercise and the slope was always constant among individuals and among subject groups. The general results concerned with sex difference in Japanese or ethnic difference between Japanese and Caucasians in both sexes could show that DL per stature was greater in males or Caucasians than in females or Japanese in young adults, however, the sex or ethnic difference disappeared in middle or old aged group. DL per alveolar volume which showed no sex or ethnic difference in young adults, was greater in middle or old aged group of females or Japanese than in that of males or Caucasians.

Adolescent↗
PubMed1989

Creep on a composite resin in water.

The compressive creep test of a composite resin (0-3.5 kg/mm2 stress levels) was conducted in water for 500 h. Linear regressions were obtained between the creep strains and the compressive stress levels at various hours. It is possible to predict the creep strain of the composite from the regression when it reaches water absorbed equilibrium after 500 h. The stress of the hygroscopic expansion was calculated from the linear regressions. The maximum stress due to the hygroscopic examination of the composite was 0.74 kg/mm2 at equilibrium of the water absorbed of the composite. The linear regressions at several compressive stress levels were obtained within 30-50 hr in the strain-log time diagrams.

Composite Resins↗
PubMed2001

The predictive parameters of erythropoietin hyporesponsiveness in patients on continuous ambulatory peritoneal dialysis.

BACKGROUND: The present study was aimed at investigating the predictive parameters of erythropoietin (epoetin) hyporesponsiveness in patients on continuous ambulatory peritoneal dialysis (CAPD). METHODS: We studied 40 patients with end-stage renal disease who had been receiving CAPD for at least 6 months and epoetin therapy for at least more than 2 months. Pearson's simple correlation and multiple stepwise linear regression analysis was used to discover what parameter can predict epoetin resistance. We expressed epoetin resistance index (ERI) as weekly epoetin dose/hematocrit/body weight'. The dose of epoetin is titrated by about 25% every 2 to 4 weeks to maintain a target hematocrit level between 33% and 36%. RESULTS: We analyzed the relationship between ERI and other predictive parameters by Pearson's correlation. These results showed ERI has a statistically significant correlation with transferrin saturation (TS) (r = -0.327, p = 0.042), total weekly Kt/Vurea (r = -0.423, p = 0.018), serum albumin level (r = -0.458, p = 0.003), normalized protein catabolic rate (nPCR) (r = -0.479, p = 0.006), normalized protein equivalent of total nitrogen appearance (nPNA) (r = -0.488, p = 0.005) and serum C-reactive protein (CRP) (r = 0.332, p = 0.036). Regression analysis was performed using stepwise linear regression for multiple variables to discover the most independent variable which is correlated with ERI. ERI was entered as a dependent variable, whereas the other parameters (age, duration of peritoneal dialysis, serum albumin level, CRP, serum ferritin, total weekly Kt/Vurea, nPCR, nPNA, serum iPTH, serum aluminium, TS) were entered as independent variables. This analysis showed CRP is the most significant variable and, if CRP is excluded, nPNA is the significant variable. CRP has a statistically significant correlation with serum albumin level (r = -0.418, p = 0.007) and total weekly Kt/Vurea (r = -0.366, p = 0.043). High CRP group has more increased level of ERI (p < 0.05), age (p < 0.05) and serum creatinine level (p < 0.05) than normal control, but more decreased level of serum albumin (p < 0.01) and serum iron levels (p < 0.05). CONCLUSION: These results indicate that CRP is the most important predictor of epoetin hyporesponsiveness.

Adult↗
PubMed2006

A comparison between three rating scales for perceived exertion and two different work tests.

In the present article, three scales developed by Borg are compared on bicycle ergometer work. In the first study, comparing the Borg Ratings of Perceived Exertion (RPE) and Category scales with Ratio properties (CR10) scales, 40 healthy subjects (12 men and eight women for each scale) with a mean age of about 30 years (SD approximately 6) participated. A work-test protocol with step-wise increase of work loads every minute was used (20 W increase for men and 15 W for women). Ratings and heart rates (HRs) were recorded every minute and blood lactates every third minute. Data obtained with the RPE scale were described with linear regressions, with individual correlations of about 0.98. Data obtained with the CR10 scale could also be described by linear regressions, but when described by power functions gave exponents of about 1.2 (SD approximately 0.4) (with one additional constant included in the power function). This was significantly lower than the exponent of between 1.5 and 1.9 that has previously been observed. Mean individual correlations were 0.98. Blood lactate concentration grew with monotonously increasing functions that could be described by power functions with a mean exponent of about 2.6 (SD approximately 0.6) (with two additional constants included in the power functions). In the second study, where also the more recently developed Borg CR100 scale (centiMax) was included, 24 healthy subjects (12 men and 12 women) with a mean age of about 29 years (SD approximately 3) participated in a work test with a step-wise increase of work loads (25 W) every third minute. Ratings and HRs were recorded. RPE values were described by linear regressions with individual correlations of about 0.97. Data from the two CR scales were described by power functions with mean exponents of about 1.4 (SD approximately 0.5) (with a-values in the power functions). Mean individual correlations were about 0.98. In both studies, a tendency for a deviation from linearity between RPE values and HRs was observed. The obtained deviations from what has previously been obtained for work of longer duration (4-6 min) points to a need for standardization of work-test protocols and to the advantage of using CR scales.

Adult↗
PubMed2005

Prediction methods for babies' birth weight using linear and nonlinear regression analysis.

The aim of this study is to determine more accurate prediction methods between linear and non-linear methods for prediction of babies' birth weight among maternal demographic characteristics. Three hundred pregnant women were included in the study. Blood glucose level before and after ingestion of glucose load, age, body mass index, % of change in weight during pregnancy, height, gestational age, parity, and fetal sex were collected as independent variables and baby birth weight as dependent variable. In linear regression, least squares estimation method was used to estimate parameters. Non-linear regression method was performed using neural network model with multilayer perceptrons, back propagation method was preferred as learning algorithm. Coefficient of determination, R2, of the linear regression equation was found 59.8% and the standard error of the estimate was calculated as 325.69 gr. In non-linear regression method R2 value was also found 59.8% and standard error of estimate was calculated as 320.30 gr. According to the results of the present study, one method is not significantly better than the other. When "accuracy in prediction" is aimed, it is better to use the two methods and compare the results, and then decide on the selection of the favourable method.

Algorithms↗