Markov blanket-based approach for learning multi-dimensional Bayesian network classifiers: An application to predict the European Quality of Life-5 Dimensions (EQ-5D) from the 39-item Parkinson's Disease Questionnaire (PDQ-39)

Multi-dimensional Bayesian network classifiers (MBCs) are probabilistic graphical models recently proposed to deal with multi-dimensional classification problems, where each instance in the data set has to be assigned to more than one class variable. In this paper, we propose a Markov blanket-based approach for learning MBCs from data. Basically, it consists of determining the Markov blanket around each class variable using the HITON algorithm, then specifying the directionality over the MBC subgraphs. Our approach is applied to the prediction problem of the European Quality of Life-5 Dimensions (EQ-5D) from the 39-item Parkinson’s Disease Questionnaire (PDQ-39) in order to estimate the health-related quality of life of Parkinson’s patients. Fivefold cross-validation experiments were carried out on randomly generated synthetic data sets, Yeast data set, as well as on a real-world Parkinson’s disease data set containing 488 patients. The experimental study, including comparison with additional Bayesian network-based approaches, back propagation for multi-label learning, multi-label k-nearest neighbor, multinomial logistic regression, ordinary least squares, and censored least absolute deviations, shows encouraging results in terms of predictive accuracy as well as the identification of dependence relationships among class and feature variables.

Regularized logistic regression and multi-objective variable selection for classifying MEG data

This paper addresses the question of maximizing classifier accuracy for classifying task-related mental activity from Magnetoencelophalography (MEG) data. We propose the use of different sources of information and introduce an automatic channel selection procedure. To determine an informative set of channels, our approach combines a variety of machine learning algorithms: feature subset selection methods, classifiers based on regularized logistic regression, information fusion, and multiobjective optimization based on probabilistic modeling of the search space. The experimental results show that our proposal is able to improve classification accuracy compared to approaches whose classifiers use only one type of MEG information or for which the set of channels is fixed a priori.

A multi-criteria optimization and decision-making approach for improvement of food engineering processes.

The objective of this study was to propose a multi-criteria optimization and decision-making technique to solve food engineering problems. This technique was demostrated using experimental data obtained on osmotic dehydratation of carrot cubes in a sodium chloride solution. The Aggregating Functions Approach, the Adaptive Random Search Algorithm, and the Penalty Functions Approach were used in this study to compute the initial set of non-dominated or Pareto-optimal solutions. Multiple non-linear regression analysis was performed on a set of experimental data in order to obtain particular multi-objective functions (responses), namely water loss, solute gain, rehydration ratio, three different colour criteria of rehydrated product, and sensory evaluation (organoleptic quality). Two multi-criteria decision-making approaches, the Analytic Hierarchy Process (AHP) and the Tabular Method (TM), were used simultaneously to choose the best alternative among the set of non-dominated solutions. The multi-criteria optimization and decision-making technique proposed in this study can facilitate the assessment of criteria weights, giving rise to a fairer, more consistent, and adequate final compromised solution or food process. This technique can be useful to food scientists in research and education, as well as to engineers involved in the improvement of a variety of food engineering processes.

Regression analysis of top of descent location for idle-thrust descents

In this paper, multiple regression analysis is used to model the top of descent (TOD) location of user-preferred descent trajectories computed by the flight management system (FMS) on over 1000 commercial flights into Melbourne, Australia. In addition to recording TOD, the cruise altitude, final altitude, cruise Mach, descent speed, wind, and engine type were also identified for use as the independent variables in the regression analysis. Both first-order and second-order models are considered, where cross-validation, hypothesis testing, and additional analysis are used to compare models. This identifies the models that should give the smallest errors if used to predict TOD location for new data in the future. A model that is linear in TOD altitude, final altitude, descent speed, and wind gives an estimated standard deviation of 3.9 nmi for TOD location given the trajectory parame- ters, which means about 80% of predictions would have error less than 5 nmi in absolute value. This accuracy is better than demonstrated by other ground automation predictions using kinetic models. Furthermore, this approach would enable online learning of the model. Additional data or further knowledge of algorithms is necessary to conclude definitively that no second-order terms are appropriate. Possible applications of the linear model are described, including enabling arriving aircraft to fly optimized descents computed by the FMS even in congested airspace.

Regression models with MoPs Bayesian networks

We present a model of Bayesian network for continuous variables, where densities and conditional densities are estimated with B-spline MoPs. We use a novel approach to directly obtain conditional densities estimation using B-spline properties. In particular we implement naive Bayes and wrapper variables selection. Finally we apply our techniques to the problem of predicting neurons morphological variables from electrophysiological ones.

System failure prediction through rare-events elastic-net logistic regression

Predicting failures in a distributed system based on previous events through logistic regression is a standard approach in literature. This technique is not reliable, though, in two situations: in the prediction of rare events, which do not appear in enough proportion for the algorithm to capture, and in environments where there are too many variables, as logistic regression tends to overfit on this situations; while manually selecting a subset of variables to create the model is error- prone. On this paper, we solve an industrial research case that presented this situation with a combination of elastic net logistic regression, a method that allows us to automatically select useful variables, a process of cross-validation on top of it and the application of a rare events prediction technique to reduce computation time. This process provides two layers of cross- validation that automatically obtain the optimal model complexity and the optimal mode l parameters values, while ensuring even rare events will be correctly predicted with a low amount of training instances. We tested this method against real industrial data, obtaining a total of 60 out of 80 possible models with a 90% average model accuracy.

Interleukin 12 and B7-1 costimulatory molecule expressed by an adenovirus vector act synergistically to facilitate tumor regression

Stimulation of antitumor immune mechanisms is the primary goal of cancer immunotherapy, and accumulating evidence suggests that effective alteration of the host–tumor relationship involves immunomodulating cytokines and also the presence of costimulatory molecules. To examine the antitumor effect of direct in vivo gene transfer of murine interleukin 12 (IL-12) and B7-1 into tumors, we developed an adenovirus (Ad) vector, AdIL12–B7-1, that encodes the two IL-12 subunits in early region 1 (E1) and the B7-1 gene in E3 under control of the murine cytomegalovirus promoter. This vector expressed high levels of IL-12 and B7-1 in infected murine and human cell lines and in primary murine tumor cells. In mice bearing tumors derived from a transgenic mouse mammary adenocarcinoma, a single intratumoral injection with a low dose (2.5 × 107 pfu/mouse) of AdIL12–B7-1 mediated complete regression in 70% of treated animals. By contrast, administration of a similar dose of recombinant virus encoding IL-12 or B7-1 alone resulted in only a delay in tumor growth. Interestingly, coinjection of two different viruses expressing either IL-12 or B7-1 induced complete tumor regression in only 30% of animals treated at this dose. Significantly, cured animals remained tumor free after rechallenge with fresh tumor cells, suggesting that protective immunity had been induced by treatment with AdIL12–B7-1. These results support the use of Ad vectors as a highly efficient delivery system for synergistically acting molecules and show that the combination of IL-12 and B7-1 within a single Ad vector might be a promising approach for in vivo cancer therapy.

Cytokine-mediated enhancement of epidermal growth factor receptor expression provides an immunological approach to the therapy of pancreatic cancer

The increased expression of epidermal growth factor receptor induced by tumor necrosis factor α renders pancreatic cancer cells more susceptible to antibody-dependent cellular cytotoxicity by a mAb specific for this receptor. Laboratory studies with athymic mice bearing xenografts of human pancreatic cancer cells demonstrated a cytokine-induced ability of the mAb to cause significant tumor regression. In a phase I/II clinical trial, 26 patients with unresectable pancreatic cancer were enrolled into three cohorts receiving variable amounts of the antibody together with a constant amount of tumor necrosis factor α. With increasing doses of antibody, the growth of the primary tumor was significantly inhibited. This was reflected by a longer median survival, with one complete remission lasting for 3 years obtained with the highest dose of antibody employed. Thus, a combination of the cytokine, tumor necrosis factor α, with a mAb to the epidermal growth factor receptor offers a potentially useful approach for the treatment of pancreatic cancer.

Gene gun-mediated skin transfection with interleukin 12 gene results in regression of established primary and metastatic murine tumors.

Particle-mediated (gene gun) in vivo delivery of the murine interleukin 12 (IL-12) gene in an expression plasmid was evaluated for antitumor activity. Transfer of IL-12 cDNA into epidermal cells overlying an implanted intradermal tumor resulted in detectable levels (266.0 +/- 27.8 pg) of the transgenic protein at the skin tissue treatment site. Despite these low levels of transgenic IL-12, complete regression of established tumors (0.4-0.8 cm in diameter) was achieved in mice bearing Renca, MethA, SA-1, or L5178Y syngeneic tumors. Only one to four treatments with IL-12 cDNA-coated particles, starting on day 7 after tumor cell implantation, were required to achieve complete tumor regression. This antitumor effect was CD8+ T cell-dependent and led to the generation of tumor-specific immunological memory. By using a metastatic P815 tumor model, we further showed that a delivery of IL-12 cDNA into the skin overlying an advanced intradermal tumor, followed by tumor excision and three additional IL-12 gene transfections, could significantly inhibit systemic metastases, resulting in extended survival of test mice. These results suggest that gene gun-mediated in vivo delivery of IL-12 cDNA should be further developed for potential clinical testing as an approach for human cancer gene therapy.

Treatment of a human breast cancer xenograft with an adenovirus vector containing an interferon gene results in rapid regression due to viral oncolysis and gene therapy.

Treatment of a human breast cancer cell line (MDA-MB-435) in nude mice with a recombinant adenovirus containing the human interferon (IFN) consensus gene, IFN-con1 (ad5/IFN), resulted in tumor regression in 100% of the animals. Tumor regression occurred when virus was injected either within 24 hr of tumor cell implantation or with established tumors. However, regression of the tumor was also observed in controls in which either the wild-type virus or a recombinant virus containing the luciferase gene was used, although tumor growth was not completely suppressed. Tumor regression was accompanied by a decrease in p53 expression. Two other tumors, the human myelogenous leukemic cell line K562 and the hamster melanoma tumor RPMI 1846, also responded to treatment but only with ad5/IFN. In the case of K562 tumors, there was complete regression of the tumor, and tumors derived from RPMI 1846 showed partial regression. We propose that the complete regression of the breast cancer with the recombinant virus ad5/IFN was the result of two events: viral oncolysis in which tumor cells are being selectively lysed by the replication-competent virus and the enhanced effect of expression of the IFN-con1 gene. K562 and RPMI 1846 tumors regressed only as a result of IFN gene therapy. This was confirmed by in vitro analysis. Our results indicate that a combination of viral oncolysis with a virus of low pathogenicity, itself resistant to the effects of IFN and IFN gene therapy, might be a fruitful approach to the treatment of a variety of different tumors, in particular breast cancers.

A composite DOP approach to excluding bottom reflectance in mapping water parameters of shallow coastal zones from TM imagery

In this paper we proposed a composite depth of penetration (DOP) approach to excluding bottom reflectance in mapping water quality parameters from Landsat thematic mapper (TM) data in the shallow coastal zone of Moreton Bay, Queensland, Australia. Three DOPs were calculated from TM1, TM2 and TM3, in conjunction with bathymetric data, at an accuracy ranging from +/-5% to +/-23%. These depths were used to segment the image into four DOP zones. Sixteen in situ water samples were collected concurrently with the recording of the satellite image. These samples were used to establish regression models for total suspended sediment (TSS) concentration and Secchi depth with respect to a particular DOP zone. Containing identical bands and their transformations for both parameters, the models are linear for TSS concentration, logarithmic for Secchi depth. Based on these models, TSS concentration and Secchi depth were mapped from the satellite image in respective DOP zones. Their mapped patterns are consistent with the in situ observed ones. Spatially, overestimation and underestimation of the parameters are restricted to localised areas but related to the absolute value of the parameters. The mapping was accomplished more accurately using multiple DOP zones than using a single zone in shallower areas. The composite DOP approach enables the mapping to be extended to areas as shallow as <3 m. (C) 2004 Elsevier Inc. All rights reserved.

Synovial sarcoma - Towards a simplified approach to prognosis

Background: Synovial sarcoma is a high grade sarcoma that usually occurs in adults. Numerous studies have attempted to identify prognostic factors that might allow more effective treatment for particular subgroups of patients. Methods: We studied 25 histologically confirmed cases of synovial sarcoma in an attempt to identify particular patient, tumour or treatment characteristics that might have a prognostic significance using Cox proportional hazards regression modelling to identify differences in survival rates. All patients received their definitive surgical treatment from a single orthopaedic surgeon reducing the likelihood of bias related to variations in surgical technique. Results: Statistically significant higher survival rates were seen in female patients (P = 0.040) and in patients aged

Assessing the impacts of grazing levels on bird density in woodland habitat: a Bayesian approach using expert opinion

Many studies on birds focus on the collection of data through an experimental design, suitable for investigation in a classical analysis of variance (ANOVA) framework. Although many findings are confirmed by one or more experts, expert information is rarely used in conjunction with the survey data to enhance the explanatory and predictive power of the model. We explore this neglected aspect of ecological modelling through a study on Australian woodland birds, focusing on the potential impact of different intensities of commercial cattle grazing on bird density in woodland habitat. We examine a number of Bayesian hierarchical random effects models, which cater for overdispersion and a high frequency of zeros in the data using WinBUGS and explore the variation between and within different grazing regimes and species. The impact and value of expert information is investigated through the inclusion of priors that reflect the experience of 20 experts in the field of bird responses to disturbance. Results indicate that expert information moderates the survey data, especially in situations where there are little or no data. When experts agreed, credible intervals for predictions were tightened considerably. When experts failed to agree, results were similar to those evaluated in the absence of expert information. Overall, we found that without expert opinion our knowledge was quite weak. The fact that the survey data is quite consistent, in general, with expert opinion shows that we do know something about birds and grazing and we could learn a lot faster if we used this approach more in ecology, where data are scarce. Copyright (c) 2005 John Wiley & Sons, Ltd.

An Incremental EM-based Learning Approach for On-Line Prediction of Hospital Resource Utilization

Objective: Inpatient length of stay (LOS) is an important measure of hospital activity, health care resource consumption, and patient acuity. This research work aims at developing an incremental expectation maximization (EM) based learning approach on mixture of experts (ME) system for on-line prediction of LOS. The use of a batchmode learning process in most existing artificial neural networks to predict LOS is unrealistic, as the data become available over time and their pattern change dynamically. In contrast, an on-line process is capable of providing an output whenever a new datum becomes available. This on-the-spot information is therefore more useful and practical for making decisions, especially when one deals with a tremendous amount of data. Methods and material: The proposed approach is illustrated using a real example of gastroenteritis LOS data. The data set was extracted from a retrospective cohort study on all infants born in 1995-1997 and their subsequent admissions for gastroenteritis. The total number of admissions in this data set was n = 692. Linked hospitalization records of the cohort were retrieved retrospectively to derive the outcome measure, patient demographics, and associated co-morbidities information. A comparative study of the incremental learning and the batch-mode learning algorithms is considered. The performances of the learning algorithms are compared based on the mean absolute difference (MAD) between the predictions and the actual LOS, and the proportion of predictions with MAD < 1 day (Prop(MAD < 1)). The significance of the comparison is assessed through a regression analysis. Results: The incremental learning algorithm provides better on-line prediction of LOS when the system has gained sufficient training from more examples (MAD = 1.77 days and Prop(MAD < 1) = 54.3%), compared to that using the batch-mode learning. The regression analysis indicates a significant decrease of MAD (p-value = 0.063) and a significant (p-value = 0.044) increase of Prop(MAD

Multi-level zero-inflated Poisson regression modelling of correlated count data with excess zeros

Count data with excess zeros relative to a Poisson distribution are common in many biomedical applications. A popular approach to the analysis of such data is to use a zero-inflated Poisson (ZIP) regression model. Often, because of the hierarchical Study design or the data collection procedure, zero-inflation and lack of independence may occur simultaneously, which tender the standard ZIP model inadequate. To account for the preponderance of zero counts and the inherent correlation of observations, a class of multi-level ZIP regression model with random effects is presented. Model fitting is facilitated using an expectation-maximization algorithm, whereas variance components are estimated via residual maximum likelihood estimating equations. A score test for zero-inflation is also presented. The multi-level ZIP model is then generalized to cope with a more complex correlation structure. Application to the analysis of correlated count data from a longitudinal infant feeding study illustrates the usefulness of the approach.