| 1. |
- Antoch, Jaromír, et al.
(författare)
-
Recursive robust regression computational aspects and comparison
- 1995
-
Ingår i: Computational Statistics & Data Analysis. - 0167-9473 .- 1872-7352. ; 19:2, s. 115-128
-
Tidskriftsartikel (refereegranskat)abstract
- The main objective of this paper is to show how algorithms for classical robust regression can be modified for recursive evaluation. It is shown that such a modification can be utilized to increase the algorithmic efficiency for convex object functions. However, for the non-convex ones it is demonstrated that recursion gives little help to find the optimal solution.
|
|
| 2. |
|
|
| 3. |
- Arslan, O., et al.
(författare)
-
Algorithms to compute CM - and S-estimates for regression
- 2003
-
Ingår i: International Conference on Robust Statistics. - : Physica-Verlag Rudolf Liebig GmbH. - 3790815187 ; , s. 62-76
-
Konferensbidrag (refereegranskat)abstract
- Constrained M-estimators for regression were introduced by Mendes and Tyler in 1995 as an alternative class of robust regression estimators with high breakdown point and high asymptotic efficiency. To compute the CM-estimate, the global minimum of an objective function with an inequality constraint has to be localized. To find the S-estimate for the same problem, we instead restrict ourselves to the boundary of the feasible region. The algorithm presented for computing CM-estimates can easily be modified to compute S-estimates as well. Testing is carried out with a comparison to the algorithm SURREAL by Ruppert
|
|
| 4. |
- Bergström, Göran, 1964, et al.
(författare)
-
Prevalence of Subclinical Coronary Artery Atherosclerosis in the General Population
- 2021
-
Ingår i: Circulation. - Philadelphia : American Heart Association. - 0009-7322 .- 1524-4539. ; 144:12, s. 916-929
-
Tidskriftsartikel (refereegranskat)abstract
- Background: Early detection of coronary atherosclerosis using coronary computed tomography angiography (CCTA), in addition to coronary artery calcification (CAC) scoring, may help inform prevention strategies. We used CCTA to determine the prevalence, severity, and characteristics of coronary atherosclerosis and its association with CAC scores in a general population.Methods: We recruited 30 154 randomly invited individuals age 50 to 64 years to SCAPIS (the Swedish Cardiopulmonary Bioimage Study). The study includes individuals without known coronary heart disease (ie, no previous myocardial infarctions or cardiac procedures) and with high-quality results from CCTA and CAC imaging performed using dedicated dual-source CT scanners. Noncontrast images were scored for CAC. CCTA images were visually read and scored for coronary atherosclerosis per segment (defined as no atherosclerosis, 1% to 49% stenosis, or ≥50% stenosis). External validity of prevalence estimates was evaluated using inverse probability for participation weighting and Swedish register data.Results: In total, 25 182 individuals without known coronary heart disease were included (50.6% women). Any CCTA-detected atherosclerosis was found in 42.1%; any significant stenosis (≥50%) in 5.2%; left main, proximal left anterior descending artery, or 3-vessel disease in 1.9%; and any noncalcified plaques in 8.3% of this population. Onset of atherosclerosis was delayed on average by 10 years in women. Atherosclerosis was more prevalent in older individuals and predominantly found in the proximal left anterior descending artery. Prevalence of CCTA-detected atherosclerosis increased with increasing CAC scores. Among those with a CAC score >400, all had atherosclerosis and 45.7% had significant stenosis. In those with 0 CAC, 5.5% had atherosclerosis and 0.4% had significant stenosis. In participants with 0 CAC and intermediate 10-year risk of atherosclerotic cardiovascular disease according to the pooled cohort equation, 9.2% had CCTA-verified atherosclerosis. Prevalence estimates had excellent external validity and changed marginally when adjusted to the age-matched Swedish background population.Conclusions: Using CCTA in a large, random sample of the general population without established disease, we showed that silent coronary atherosclerosis is common in this population. High CAC scores convey a significant probability of substantial stenosis, and 0 CAC does not exclude atherosclerosis, particularly in those at higher baseline risk.
|
|
| 5. |
- Bergström, Göran, et al.
(författare)
-
Prevalence of Subclinical Coronary Artery Atherosclerosis in the General Population
- 2021
-
Ingår i: Circulation. - : Wolters Kluwer. - 0009-7322 .- 1524-4539. ; 144:12, s. 916-929
-
Tidskriftsartikel (refereegranskat)abstract
- Background: Early detection of coronary atherosclerosis using coronary computed tomography angiography (CCTA), in addition to coronary artery calcification (CAC) scoring, may help inform prevention strategies. We used CCTA to determine the prevalence, severity, and characteristics of coronary atherosclerosis and its association with CAC scores in a general population.Methods: We recruited 30 154 randomly invited individuals age 50 to 64 years to SCAPIS (the Swedish Cardiopulmonary Bioimage Study). The study includes individuals without known coronary heart disease (ie, no previous myocardial infarctions or cardiac procedures) and with high-quality results from CCTA and CAC imaging performed using dedicated dual-source CT scanners. Noncontrast images were scored for CAC. CCTA images were visually read and scored for coronary atherosclerosis per segment (defined as no atherosclerosis, 1% to 49% stenosis, or ≥50% stenosis). External validity of prevalence estimates was evaluated using inverse probability for participation weighting and Swedish register data.Results: In total, 25 182 individuals without known coronary heart disease were included (50.6% women). Any CCTA-detected atherosclerosis was found in 42.1%; any significant stenosis (≥50%) in 5.2%; left main, proximal left anterior descending artery, or 3-vessel disease in 1.9%; and any noncalcified plaques in 8.3% of this population. Onset of atherosclerosis was delayed on average by 10 years in women. Atherosclerosis was more prevalent in older individuals and predominantly found in the proximal left anterior descending artery. Prevalence of CCTA-detected atherosclerosis increased with increasing CAC scores. Among those with a CAC score >400, all had atherosclerosis and 45.7% had significant stenosis. In those with 0 CAC, 5.5% had atherosclerosis and 0.4% had significant stenosis. In participants with 0 CAC and intermediate 10-year risk of atherosclerotic cardiovascular disease according to the pooled cohort equation, 9.2% had CCTA-verified atherosclerosis. Prevalence estimates had excellent external validity and changed marginally when adjusted to the age-matched Swedish background population.Conclusions: Using CCTA in a large, random sample of the general population without established disease, we showed that silent coronary atherosclerosis is common in this population. High CAC scores convey a significant probability of substantial stenosis, and 0 CAC does not exclude atherosclerosis, particularly in those at higher baseline risk.
|
|
| 6. |
|
|
| 7. |
- Brandell, Gerd, et al.
(författare)
-
Encouraging More Women Into Computer Science : Initiating a Single-Sex Intervention Program in Sweden
- 1997
-
Ingår i: Mathematics Education Research Journal. - 1033-2170 .- 2211-050X. ; 9:3, s. 312-324
-
Tidskriftsartikel (refereegranskat)abstract
- The process of starting a new program in computer science and engineering, heavily based on applied mathematics and only open to women, is described in this paper. The program was introduced into an educational system without any tradition in single-sex education. Important observations made during the process included the considerable interest in mathematics and curiosity about computer science found among female students at the secondary school level, and the acceptance of the single-sex program by the staff, administration, and management of the university as well as among male and female students. The process described highlights the importance of preparing the environment for a totally new type of educational program.
|
|
| 8. |
|
|
| 9. |
|
|
| 10. |
- Edlund, Ove, et al.
(författare)
-
Algorithms for non-linear M-estimation
- 1997
-
Ingår i: Computational statistics (Zeitschrift). - 0943-4062 .- 1613-9658. ; 12:3, s. 373-383
-
Tidskriftsartikel (refereegranskat)
|
|
