| 1. |
- Corander, Jukka, et al.
(author)
-
A bayesian random fragment insertion model for de novo detection of DNA regulatory binding regions
- 2007
-
Other publication (other academic/artistic)abstract
- Identification of regulatory binding motifs within DNA sequences is a commonly occurring problem in computationnl bioinformatics. A wide variety of statistical approaches have been proposed in the literature to either scan for previously known motif types or to attempt de novo identification of a fixed number (typically one) of putative motifs. Most approaches assume the existence of reliable biodatabasc information to build probabilistic a priori description of the motif classes. No method has been previously proposed for finding the number of putative de novo motif types and their positions within a set of DNA sequences. As the number of sequenced genomes from a wide variety of organisms is constantly increasing, there is a clear need for such methods. Here we introduce a Bayesian unsupervised approach for this purpose by using recent advances in the theory of predictive classification and Markov chain Monte Carlo computation. Our modelling framework enables formal statistical inference in a large-scale sequence screening and we illustrate it by a set of examples.
|
|
| 2. |
- Corander, Jukka, et al.
(author)
-
Parallell interacting MCMC for learning of topologies of graphical models
- 2008
-
In: Data mining and knowledge discovery. - : Springer Science and Business Media LLC. - 1384-5810 .- 1573-756X. ; 17:3, s. 431-456
-
Journal article (peer-reviewed)abstract
- Automated statistical learning of graphical models from data has attained a considerable degree of interest in the machine learning and related literature. Many authors have discussed and/or demonstrated the need for consistent stochastic search methods that would not be as prone to yield locally optimal model structures as simple greedy methods. However, at the same time most of the stochastic search methods are based on a standard Metropolis-Hastings theory that necessitates the use of relatively simple random proposals and prevents the utilization of intelligent and efficient search operators. Here we derive an algorithm for learning topologies of graphical models from samples of a finite set of discrete variables by utilizing and further enhancing a recently introduced theory for non-reversible parallel interacting Markov chain Monte Carlo-style computation. In particular, we illustrate how the non-reversible approach allows for novel type of creativity in the design of search operators. Also, the parallel aspect of our method illustrates well the advantages of the adaptive nature of search operators to avoid trapping states in the vicinity of locally optimal network topologies.
|
|
| 3. |
- Ekdahl, Anne W, et al.
(author)
-
Costs and Effects of an Ambulatory Geriatric Unit (the AGe-FIT Study) : A Randomized Controlled Trial
- 2015
-
In: Journal of the American Medical Directors Association. - : Elsevier. - 1538-9375 .- 1525-8610. ; 16:6, s. 497-503
-
Journal article (peer-reviewed)abstract
- OBJECTIVES: To examine costs and effects of care based on comprehensive geriatric assessment (CGA) provided by an ambulatory geriatric care unit (AGU) in addition to usual care.DESIGN: Assessor-blinded, single-center randomized controlled trial.SETTING: AGU in an acute hospital in southeastern Sweden.PARTICIPANTS: Community-dwelling individuals aged 75 years or older who had received inpatient hospital care 3 or more times in the past 12 months and had 3 or more concomitant medical diagnoses were eligible for study inclusion and randomized to the intervention group (IG; n = 208) or control group (CG; n = 174). Mean age (SD) was 82.5 (4.9) years.INTERVENTION: Participants in the IG received CGA-based care at the AGU in addition to usual care.OUTCOME MEASURES: The primary outcome was number of hospitalizations. Secondary outcomes were days in hospital and nursing home, mortality, cost of public health and social care, participant' sense of security in care, and health-related quality of life (HRQoL).RESULTS: Baseline characteristics did not differ between groups. The number of hospitalizations did not differ between the IG (2.1) and CG (2.4), but the number of inpatient days was lower in the IG (11.1 vs 15.2; P = .035). The IG showed trends of reduced mortality (hazard ratio 1.51; 95% confidence interval [CI] 0.988-2.310; P = .057) and an increased sense of security in care interaction. No difference in HRQoL was observed. Costs for the IG and CG were 33,371£ (39,947£) and 30,490£ (31,568£; P = .432).CONCLUSIONS AND RELEVANCE: This study of CGA-based care was performed in an ambulatory care setting, in contrast to the greater part of studies of the effects of CGA, which have been conducted in hospital settings. This study confirms the superiority of this type of care to elderly people in terms of days in hospital and sense of security in care interaction and that a shift to more accessible care for older people with multimorbidity is possible without increasing costs. This study can aid the planning of future interventions for older people.TRIAL REGISTRATION: clinicaltrials.gov identifier: NCT01446757.
|
|
| 4. |
- Ekdahl, Anne Wissendorff, et al.
(author)
-
Is care based on comprehensive geriatric assessment with mobile teams better than usual care? : A study protocol of a randomised controlled trial (The GerMoT study)
- 2018
-
In: BMJ Open. - : BMJ. - 2044-6055. ; 8:10
-
Journal article (peer-reviewed)abstract
- INTRODUCTION: Comprehensive geriatric assessment (CGA) is a multidimensional, interdisciplinary diagnostic process used to determine the medical, psychological and functional capabilities of frail older people. The primary aim of our current study is to confirm whether CGA-based outpatient care is superior than usual care in terms of health-related outcomes, resource use and costs.METHODS AND ANALYSIS: The Geriatric Mobile Team trial is designed as a single-centre randomised, controlled, assessor-blinded (at baseline) trial. All participants will be identified via local healthcare registries with the following inclusion criteria: age ≥75 years, ≥3 different diagnoses and ≥3 visits to the emergency care unit (with or without admittance to hospital) during the past 18 months. Nursing home residency will be an exclusion criterion. Baseline assessments will be done before the 1:1 randomisation. Participants in the intervention group will, after an initial CGA, have access to care given by a geriatric team in addition to usual care. The control group receives usual care only. The primary outcome is the total number of inpatient days during the follow-up period. Assessments of the outcomes: mortality, quality of life, health care use, physical functional level, frailty, dependence and cognition will be performed 12 and 24 months after inclusion. Both descriptive and analytical statistics will be used, in order to compare groups and for analyses of outcomes over time including changes therein. The primary outcome will be analysed using analysis of variance, including in-transformed values if needed to achieve normal distribution of the residuals.ETHICS AND DISSEMINATION: Ethical approval has been obtained and the results will be disseminated in national and international journals and to health care leaders and stakeholders. Protocol amendments will be published in ClinicalTrials.gov as amendments to the initial registration NCT02923843. In case of success, the study will promote the implementation of CGA in outpatient care settings and thereby contribute to an improved care of older people with multimorbidity through dissemination of the results through scientific articles, information to politicians and to the public.TRIAL REGISTRATION NUMBER: NCT02923843; Pre-results.
|
|
| 5. |
- Ekdahl, Anne W., et al.
(author)
-
Long-Term Evaluation of the Ambulatory Geriatric Assessment: A Frailty Intervention Trial (AGe-FIT): Clinical Outcomes and Total Costs After 36 Months
- 2016
-
In: Journal of the American Medical Directors Association. - : ELSEVIER SCIENCE INC. - 1525-8610 .- 1538-9375. ; 17:3, s. 263-268
-
Journal article (peer-reviewed)abstract
- Objective: To compare the effects of care based on comprehensive geriatric assessment (CGA) as a complement to usual care in an outpatient setting with those of usual care alone. The assessment was performed 36 months after study inclusion. Design: Randomized, controlled, assessor-blinded, single-center trial. Setting: A geriatric ambulatory unit in a municipality in the southeast of Sweden. Participants: Community-dwelling individuals aged >= 75 years who had received inpatient hospital care 3 or more times in the past 12 months and had 3 or more concomitant medical diagnoses were eligible for study inclusion. Participants were randomized to the intervention group (IG) or control group (CG). Intervention: Participants in the IG received CGA-based care for 24 to 31 months at the geriatric ambulatory unit in addition to usual care. Outcome measures: Mortality, transfer to nursing home, days in hospital, and total costs of health and social care after 36 months. Results: Mean age (SD) of participants was 82.5 (4.9) years. Participants in the IG (n = 208) lived 69 days longer than did those in the CG (n = 174); 27.9% (n = 58) of participants in the IG and 38.5% (n = 67) in the CG died (hazard ratio 1.49, 95% confidence interval 1.05-2.12, P =.026). The mean number of inpatient days was lower in the IG (15.1 [SD 18.4]) than in the CG (21.0 [SD 25.0], P =.01). Mean overall costs during the 36-month period did not differ between the IG and CG (USD 71,905 [SD 85,560] and USD 65,626 [SD 66,338], P =.43). Conclusions: CGA-based care resulted in longer survival and fewer days in hospital, without significantly higher cost, at 3 years after baseline. These findings add to the evidence of CGAs superiority over usual care in outpatient settings. As CGA-based care leads to important positive outcomes, this method should be used more extensively in the treatment of older people to meet their needs. (c) 2016 AMDA - The Society for Post-Acute and Long-Term Care Medicine.
|
|
| 6. |
- Ekdahl, Magnus, 1979-
(author)
-
Approximations of Bayes Classifiers for Statistical Learning of Clusters
- 2006
-
Licentiate thesis (other academic/artistic)abstract
- It is rarely possible to use an optimal classifier. Often the classifier used for a specific problem is an approximation of the optimal classifier. Methods are presented for evaluating the performance of an approximation in the model class of Bayesian Networks. Specifically for the approximation of class conditional independence a bound for the performance is sharpened.The class conditional independence approximation is connected to the minimum description length principle (MDL), which is connected to Jeffreys’ prior through commonly used assumptions. One algorithm for unsupervised classification is presented and compared against other unsupervised classifiers on three data sets.
|
|
| 7. |
- Ekdahl, Magnus, et al.
(author)
-
Bounds for the loss in probability of correct classification under model based approximation
- 2006
-
In: Journal of machine learning research. - 1532-4435 .- 1533-7928. ; 7, s. 2449-2480
-
Journal article (peer-reviewed)abstract
- In many pattern recognition/classification problem the true class conditional model and class probabilities are approximated for reasons of reducing complexity and/or of statistical estimation. The approximated classifier is expected to have worse performance, here measured by the probability of correct classification. We present an analysis valid in general, and easily computable formulas for estimating the degradation in probability of correct classification when compared to the optimal classifier. An example of an approximation is the Naive Bayes classifier. We show that the performance of the Naive Bayes depends on the degree of functional dependence between the features and labels. We provide a sufficient condition for zero loss of performance, too.
|
|
| 8. |
- Ekdahl, Magnus, et al.
(author)
-
Concentrated or non-concentrated discrete distributions are almost independent
- 2007
-
Other publication (other academic/artistic)abstract
- The task of approximating a simultaneous distribution with a product of distributions in a single variable is important in the theory and applications of classification and learning, probabilistic reasoning, and random algmithms. The evaluation of the goodness of this approximation by statistical independence amounts to bounding uniformly upwards the difference between a joint distribution and the product of the distributions (marginals). In this paper we develop a bound that uses information about the most probable state to find a sharp estimate, which is often as sharp as possible. We also examine the extreme cases of concentration and non-conccntmtion, respectively, of the approximated distribution.
|
|
| 9. |
- Ekdahl, Magnus, 1979-
(author)
-
On approximations and computations in probabilistic classification and in learning of graphical models
- 2007
-
Doctoral thesis (other academic/artistic)abstract
- Model based probabilistic classification is heavily used in data mining and machine learning. For computational learning these models may need approximation steps however. One popular approximation in classification is to model the class conditional densities by factorization, which in the independence case is usually called the ’Naïve Bayes’ classifier. In general probabilistic independence cannot model all distributions exactly, and not much has been published on how much a discrete distribution can differ from the independence assumption. In this dissertation the approximation quality of factorizations is analyzed in two articles.A specific class of factorizations is the factorizations represented by graphical models. Several challenges arise from the use of statistical methods for learning graphical models from data. Examples of problems include the increase in the number of graphical model structures as a function of the number of nodes, and the equivalence of statistical models determined by different graphical models. In one article an algorithm for learning graphical models is presented. In the final article an algorithm for clustering parts of DNA strings is developed, and a graphical representation for the remaining DNA part is learned.
|
|
| 10. |
- Ekdahl, Magnus, 1979-, et al.
(author)
-
On Concentration of Discrete Distributions with Applications to Supervised Learning of Classifiers
- 2007
-
In: Machine Learning and Data Mining in Pattern Recognition. - Berlin, Heidelberg : Springer Berlin/Heidelberg. - 9783540734987 - 9783540734994 - 3540734988 ; , s. 2-16
-
Book chapter (peer-reviewed)abstract
- Computational procedures using independence assumptions in various forms are popular in machine learning, although checks on empirical data have given inconclusive results about their impact. Some theoretical understanding of when they work is available, but a definite answer seems to be lacking. This paper derives distributions that maximizes the statewise difference to the respective product of marginals. These distributions are, in a sense the worst distribution for predicting an outcome of the data generating mechanism by independence. We also restrict the scope of new theoretical results by showing explicitly that, depending on context, independent ('Naïve') classifiers can be as bad as tossing coins. Regardless of this, independence may beat the generating model in learning supervised classification and we explicitly provide one such scenario.
|
|
