Logistic Classification Model Research Articles

Patterns and processes of population dynamics with fluctuating habitat size: a case study of a marine copepod inhabiting tide pools

The logistic model is a fundamental population model often used as the basis for analyzing wildlife population dynamics. In the classic logistic model, however, population dynamics may be difficult to characterize if habitat size is temporally variable because population density can vary at a constant abundance, which results in variable strength of density‐dependent feedback for a given population size. To incorporate habitat size variability, we developed a general population model in which changes in population abundance, density, and habitat size are taken into account. From this model, we deduced several predictions for patterns and processes of population dynamics: 1) patterns of fluctuation in population abundance and density can diverge, with respect of their correlation and relative variability; and 2) along with density dependence, habitat size fluctuation can affect population growth with a time lag because changes in habitat size result in changes in population density. In order to test these predictions, we applied our model to population dynamics data of 36 populations of Tigriopus japonicus, a marine copepod inhabiting tide pools of variable sizes caused by weather processes. As expected, we found a significant difference in the fluctuation patterns of population abundance and density of T. japonicus populations with respect to the correlation between abundance and density and their relative variability, which correlates positively with the variability of habitat size. In addition, we found direct and lagged‐indirect effects of weather processes on population growth, which were associated with density dependence and impose regulatory forces on local and regional population dynamics. These results illustrate how changes in habitat size can have an impact on patterns and processes of wildlife population dynamics. We suggest that without knowledge of habitat size fluctuation, measures of population size and its variability as well as inferences about the processes of population dynamics may be misleading.

Postharvest citrus mass and size estimation using a logistic classification model and a watershed algorithm

A machine vision system for estimation of citrus fruit mass, fruit count, and fruit size during postharvest processing was investigated with the aim of developing of an advanced citrus yield mapping system. Such yield mapping system enables the citrus growers to efficiently manage the in-grove spatial variability factors such as: soil type, soil fertility, moisture content, etc., and can help increase yield and profits. Thus, a machine vision system was developed and installed in a citrus debris cleaning machine, which removes debris from mechanically harvested loads. An image processing algorithm was developed to identify fruit from images of the postharvest citrus from a commercial citrus grove. For fruit detection, logistic regression model based pixel classification algorithms were developed. To avoid misclassification due to highly saturated area on fruit and non-fruit regions, a highly saturated area recovering (HSAR) algorithm was developed that obviated use of a “filling holes” operation. A mass calibration process was conducted, and fruit mass was estimated, which turned out to be reasonably good. The highest coefficient of determination ( R 2 ) value between the measured fruit mass and the estimated fruit mass was observed to be 0.945 and the root mean square error was 116.1 kg. A H-minima transform based watershed algorithm was used to separate the joined fruit and enabled an estimation of fruit counting and fruit size. Fruit mass estimation using the fruit size information was also conducted and these results were compared with that of the mass estimation based on fruit pixel area. ► Machine vision based post-harvest citrus mass and size estimation. ► Image processing algorithm to detect and segment fruit image using logistic classification model. ► Highly saturated area recovering algorithm to increase segmentation accuracy. ► Neighbouring fruit separation using H-minima transform based watershed transform.

Growth curve prediction from optical density data

A fundamental aspect of predictive microbiology is the shape of the microbial growth curve and many models are used to fit microbial count data, the modified Gompertz and Baranyi equation being two of the most widely used. Rapid, automated methods such as turbidimetry have been widely used to obtain growth parameters, but do not directly give the microbial growth curve. Optical density (OD) data can be used to obtain the specific growth rate and if used in conjunction with the known initial inocula, the maximum population data and knowledge of the microbial number at a predefined OD at a known time then all the information required for the reconstruction of a standard growth curve can be obtained.Using multiple initial inocula the times to detection (TTD) at a given standard OD were obtained from which the specific growth rate was calculated. The modified logistic, modified Gompertz, 3-phase linear, Baranyi and the classical logistic model (with or without lag) were fitted to the TTD data. In all cases the modified logistic and modified Gompertz failed to reproduce the observed linear plots of the log initial inocula against TTD using the known parameters (initial inoculum, MPD and growth rate). The 3 phase linear model (3PLM), Baranyi and classical logistic models fitted the observed data and were able to reproduce elements of the OD incubation-time curves. Using a calibration curve relating OD and microbial numbers, the Baranyi equation was able to reproduce OD data obtained for Listeria monocytogenes at 37 and 30°C as well as data on the effect of pH (range 7.05 to 3.46) at 30°C.The Baranyi model was found to be the most capable primary model of those examined (in the absence of lag it defaults to the classic logistic model). The results suggested that the modified logistic and the modified Gompertz models should not be used as Primary models for TTD data as they cannot reproduce the observed data.

A Comparative Study of the Use of Generalized Additive Models and Generalized Linear Models in Tourism Research

ABSTRACTIn this paper, we investigate the impact that spatial and micro‐economic variables have on the probability that a household goes on holiday. In doing so, we propose two alternative modelling specifications: a classic discrete choice model and a semiparametric logistic model. The semiparametric model extends the classic logistic model, usually employed in studies on participation in tourism, allowing modelling in a flexible manner for continuous predictors without making any a priori assumption. This is achieved via the use of penalized regression splines. A sample of Italian households was considered for our study. Comparing the results of the two approaches, we found that both methods opportunely captured, in terms of signs, the relationships under investigation. However, the use of a more flexible approach has allowed us to uncover some interesting non‐linearities that are usually not assumed a priori, thus improving the interpretation of the results. Copyright © 2011 John Wiley & Sons, Ltd.

Comparing Bayesian network, artificial neural networks, classification trees and classical logistic models in quantitative risk assessment: an application to the European Registry of Foreign Body Injuries in Children

Risk assessment is the systematic study of decisions subject to uncertain consequences.An increasing interest has been focused on modelling techniques like Bayesian networks (BNs) since their capability of (1) combining...

Patterns of post‐glacial spread and the extent of glacial refugia of European beech (Fagus sylvatica)

AbstractAim Establishing possible relationships between the magnitudes of the glacial distribution of the European beech, Fagus sylvatica L., and its post‐glacial spread.Location Europe.Methods A database of over 400 pollen records has been used to locate Fagus populations at the end of the last glacial and during the post‐glacial in Europe and to assess the areal extent of their past distribution.Results The rate of late‐glacial and post‐glacial increase in the number of pollen sites where Fagus was locally present conforms well to a logistic model of population growth. This suggests that the area occupied by beech populations expanded exponentially from the glacial refugia for a duration of over 10,000 years, until about 3500 yr bp. In the past three millennia beech populations increased at a slower rate, tending towards an equilibrium value.Main conclusions The conformity of the increase in beech distribution to the classical logistic model of population growth indicates that: (1) a multiplicative biological process was the main factor shaping the pattern of the post‐glacial expansion of F. sylvatica in Europe, (2) climate conditions, human activity and competition may have influenced its rate of spread, and (3) beech populations did not expand with a moving closed front, but with a diffuse spread from scattered nuclei. The distribution of Fagus in Europe at the end of the last glacial appears to have been of two orders of magnitude less extensive than at present. Pleistocene refugia were likely to have been a mosaic of sparse stands of small populations scattered in multiple regions. Fagus populations appear to have increased very slowly and to a moderate extent in southern Europe, where they are now declining slightly. The central European populations increased quickly and extensively, reaching northern Europe, and are now approaching their carrying capacity.

Bayesian Curve Classification Using Wavelets

We propose classification models for binary and multicategory data where the predictor is a random function. We use Bayesian modeling with wavelet basis functions that have nice approximation properties over a large class of functional spaces and can accommodate a wide variety of functional forms observed in real life applications. We develop an unified hierarchical model to encompass both the adaptive wavelet-based function estimation model and the logistic classification model. We couple together these two models are to borrow strengths from each other in a unified hierarchical framework. The use of Gibbs sampling with conjugate priors for posterior inference makes the method computationally feasible. We compare the performance of the proposed model with other classification methods, such as the existing naive plug-in methods, by analyzing simulated and real data sets.

Comparison of the classification ability of the electrocardiogram and vectorcardiogram

Controversy exists over the classification ability of the standard 12-lead electrocardiogram (EGG) and the vectorcardiogram (VCG). In this study the diagnostic information content and classification performance of the ECG and VCG were examined using multivariate statistical techniques and a large validated data base of 3,266 cases. Logistic classification models were developed to differentiate between 7 diagnostic entities: normal (n = 538), left (n = 557), right (n = 323) and biventricular (n = 437) hypertrophy, and anterior (n = 390), inferior (n = 657) and combined (n = 364) myocardial infarction. The models were obtained from a learning sample (n = 2,446) using an optimal set of computer derived ECG and VCG measurements. They were subsequently applied to a test sample (n = 820). In the learning sample, the discrimination models resulted in a total correct classification rate of 69.6% for the ECG and 69.4% for the VCG. The total accuracy rate was slightly lower in the test set: 66.3% for the ECG and 67.1% for the VCG. The combined use of the best ECG and VCG variables did not increase total diagnostic accuracy. When cases with biventricular hypertrophy and combined infarction were deleted, accuracy rates of more than 80% were achieved for both lead systems. Differences in the classification rates for the subgroups were not statistically significant. Thus, the conventional 12-lead ECG is as good as the VCG for the differential diagnosis of 7 main entities, provided identical procedures are used in the design of the classifiers.(ABSTRACT TRUNCATED AT 250 WORDS)

Validation of a model for predicting aptitude for introductory computing

A double cross-validation design was used to assess the practical predictive value of a logistic classification model developed to predict “high and low aptitude” for introductory computing. The validation study showed that the model would have a predictive accuracy of approximately 75% in actual application. The model variables were checked by formal hypotheses tests. The results of the study indicated that the classification model would be a useful tool for counseling and formation of “high and low aptitude” lecture sections in introductory computing.

Non-autonomous logistic equations as models of populations in a deteriorating environment

Abstract The non-autonomous logistic equation d x(t) d t = r(t)x(t)[1 − x(t) K(t) ] is studied under conditions that include an environment which is completely deteriorating. In this setting, when the population's growth rate, r, is large on the average, solutions track the environment with a consequent extinction of the population. However, when both r and rK−1 are small in the sense that they are in L1[0,∞) then an asymptotic equivalence, where all solutions tend to positive limits as t approaches infinity, results and the population is persistent, independent of initial density. The asymptotic equivalence produces an unreasonable overshoot of carrying capacity which leads to concern about employing the logistic equation in the above form as a population model when growth rates are close to zero. A re-interpretation of the parameters of the logistic equation leads to the alternative logistic formulation d x(t) d t = x(t)[r(t) − c B(t) x(t)], (c > 0) . A biological interpretation of the parameters is presented and this equation is compared with the classical logistic model in the case where the parameters are constant. If the alternative logistic model is applied in a situation with time-varying parameters, then a deteriorating environment always leads to extinction of the population regardless of the behavior of r. Similarly, a growth rate which is small on the average results in extinction regardless of the behavior of B. Furthermore, r and B have limiting values as t approaches infinity then so does x and the terminal value of x is equal to the terminal value of the carrying capacity of the population. In general, the alternative formulation seems to be the more reasonable model in situations where perturbations lead to severe decreases in environmental quality and growth rates.