Search results for "Random"
showing 10 items of 3931 documents
Improvements and applications of the elements of prototype-based clustering
2018
Clustering or cluster analysis is an essential part of data mining, machine learning, and pattern recognition. The most popularly applied clustering methods are partitioning-based or prototype-based methods. Prototype-based clustering methods usually have easy implementability and good scalability. These methods, such as K-means clustering, have been used for different applications in various fields. On the other hand, prototype-based clustering methods are typically sensitive to initialization, and the selection of the number of clusters for knowledge discovery purposes is not straightforward. In the era of big data, in high-velocity, ever-growing datasets, which can also be erroneous, outl…
Random utility approach to re-design passengers interchange nodes
2009
The passengers interchange node is a complex infrastructure, where the user can choose among different modal options to do his/her trip. These transport infrastructures have various terminals (car park, bus stop, metro, light rail, railway, extra-urban and sub urban bus and so on). The passengers, inside interchange node, have to reach these terminals walking often with luggage and for long distances. We calibrated a discrete choice model taking into account attributes able to explain the passenger behaviour by observed flows and generalized least squares technique. Furthermore an ex ante / ex post analysis was carried out in order to design and evaluate measure to incentive modal integrati…
On probabilistic interpretations of predicates
2016
In classical logic, any m-ary predicate is interpreted as an m-argument two-valued relation defined on a non-empty universe. In probability theory, m-ary predicates are interpreted as probability measures on the mth power of a probability space. m-ary probabilistic predicates are equivalently semantically characterized as m-dimensional cumulative distribution functions defined on Rm. The paper is mainly concerned with probabilistic interpretations of unary predicates in the algebra of cumulative distribution functions defined on R. This algebra, enriched with two constants, forms a bounded De Morgan algebra. Two logical systems based on the algebra of cumulative distributions are defined an…
Rationale and design of the DARWIN-T2D (DApagliflozin Real World evIdeNce in Type 2 Diabetes): A multicenter retrospective nationwide Italian study a…
2017
Background Randomized controlled trials (RCTs) in the field of diabetes have limitations inherent to the fact that design, setting, and patient characteristics may be poorly transferrable to clinical practice. Thus, evidence from studies using routinely accumulated clinical data are increasingly valued. Aims We herein describe rationale and design of the DARWIN-T2D (DApagliflozin Real World evIdeNce in Type 2 Diabetes), a multicenter retrospective nationwide study conducted at 50 specialist outpatient clinics in Italy and promoted by the Italian Diabetes Society. Data synthesis The primary objective of the study is to describe the baseline clinical characteristics (particularly HbA1c) of pa…
Managing for Results in Primary Education: Results of a Randomized Trial in Madagascar
2012
International audience; Using information from a randomized experiment carried out over the course of two school years in Madagascar, this paper evaluates the impact of specific actions designed to streamline and tighten the work processes of public primary school directors. The results show that interventions at the school level, reinforced by interventions at the district and subdistrict levels, succeeded overall in changing school heads' behavior toward better management. However, the average impact hides important heterogeneity. The impact of the intervention was significantly larger among school heads who had a nonpermanent contract and among school principals who were responsible for …
Nošķeltās vidējās vērtības pielietošana trūkstošu datu gadījumā randomizētos kontrolētos pētījumos
2022
Šajā darbā tiek apskatīta nošķeltā vidējā vērtība trūkstošu datu gadījumā un hipotēžu testi nošķelto vidējo vērtību salīdzināšanai -- permutāciju, Jūenas un EL tests. Tiek sniegts arī ieskats par daudzkārtējo imputāciju. Tiek izstrādāts simulāciju eksperiments, kur metodes tiek apskatītas praktiski un rezultāti tiek attēloti tabulās, lai tos būtu vieglāk salīdzināt. Tiek analizēts arī reāls datu piemērs. Darbā var novērot atšķirīgus rezultātus starp metodēm.
Meditāciju saistība ar smadzeņu darbību, izmantojot elektroencefalogrāfu (EEG), randomizēti kontrolētos pētījumos: metaanalīze
2020
Mūsdienās meditācijas tiek praktizētas ne tikai austrumos, bet kļuvušas populāras arī rietumu pasaulē. Šajā metaanalīzē tika meklēta atbilde uz jautājumu, kādas saistības pastāv starp meditācijām un galvas smadzeņu viļņu darbību pieaugušajiem randomizēti kontrolētos pētījumos. Metaanalīzē tika iekļauti trīs pētījumi ar alfas viļņu rādītājiem un kopējo izlases apjomu N = 94. Metaanalīzes aprēķini uzrāda nelielu statistiski nozīmīgu efektu r = 0,20 (p<0,001). Tādējādi šī pētījuma ietvaros secināms, ka pēc meditācijām novērojams neliels efekts uz galvas smadzeņu alfa viļņu rādītājiem. Tas saskan ar līdzšinējiem pētījumiem, kuros vairākkārt ir secināts, ka meditācijas spēj statistiski nozīmīgi …
ENSEMBLE METHODS FOR RANKING DATA
2017
The last years have seen a remarkable flowering of works about the use of decision trees for ranking data. As a matter of fact, decision trees are useful and intuitive, but they are very unstable: small perturbations bring big changes. This is the reason why it could be necessary to use more stable procedures, as ensemble methods, in order to find which predictors are able to explain the preference structure. In this work ensemble methods as BAGGING and Random Forest are proposed, from both a theoretical and computational point of view, for deriving classification trees when ranking data are observed. The advantages of these procedures are shown through an example on the SUSHI data set.
Eating behavior dimensions and 9-year weight loss maintenance : a sub-study of the Finnish Diabetes prevention study
2023
Abstract Background Behavioral processes through which lifestyle interventions influence risk factors for type 2 diabetes (T2DM), e.g., body weight, are not well-understood. We examined whether changes in psychological dimensions of eating behavior during the first year of lifestyle intervention would mediate the effects of intervention on body weight during a 9-year period. Methods Middle-aged participants (38 men, 60 women) with overweight and impaired glucose tolerance (IGT) were randomized to an intensive, individualized lifestyle intervention group (n = 51) or a control group (n = 47). At baseline and annually thereafter until nine years body weight was measured and the Three Factor Ea…
Computational aspects in checking of coherence and propagation of conditional probability bounds
2000
In this paper we consider the problem of reducing the computational difficulties in g-coherence checking and propagation of imprecise conditional probability assessments. We review some theoretical results related with the linear structure of the random gain in the betting criterion. Then, we propose a modi ed version of two existing algorithms, used for g-coherence checking and propagation, which are based on linear systems with a reduced number of unknowns. The reduction in the number of unknowns is obtained by an iterative algorithm. Finally, to illustrate our procedure we give some applications.