| 1. |
- Elden, Lars
(författare)
-
Numerical linear algebra in data mining
- 2006
-
Ingår i: Acta Numerica. - 0962-4929 .- 1474-0508. ; 15, s. 327-384
-
Forskningsöversikt (refereegranskat)abstract
- Ideas and algorithms from numerical linear algebra are important in several areas of data mining. We give an overview of linear algebra methods in text mining (information retrieval), pattern recognition (classification of handwritten digits), and Page Rank computations for web search engines. The emphasis is on rank reduction as a method of extracting information from a data matrix, low-rank approximation of matrices using the singular value decomposition and clustering, and on eigenvalue methods for network analysis. © Cambridge University Press, 2006.
|
|
| 3. |
- Chaturvedi, Nish, et al.
(författare)
-
Effect of candesartan on prevention (DIRECT-Prevent 1) and progression (DIRECT-Protect 1) of retinopathy in type 1 diabetes: randomised, placebo-controlled trials.
- 2008
-
Ingår i: Lancet. - 1474-547X. ; 372:9647, s. 1394-402
-
Tidskriftsartikel (refereegranskat)abstract
- BACKGROUND: Results of previous studies suggest that renin-angiotensin system blockers might reduce the burden of diabetic retinopathy. We therefore designed the DIabetic REtinopathy Candesartan Trials (DIRECT) Programme to assess whether candesartan could reduce the incidence and progression of retinopathy in type 1 diabetes. METHODS: Two randomised, double-blind, parallel-design, placebo-controlled trials were done in 309 centres worldwide. Participants with normotensive, normoalbuminuric type 1 diabetes without retinopathy were recruited to the DIRECT-Prevent 1 trial and those with existing retinopathy were recruited to DIRECT-Protect 1, and were assigned to candesartan 16 mg once a day or matching placebo. After 1 month, the dose was doubled to 32 mg. Investigators and participants were unaware of the treatment allocation status. The primary endpoints were incidence and progression of retinopathy and were defined as at least a two-step and at least a three-step increase on the Early Treatment Diabetic Retinopathy Study (ETDRS) scale, respectively. These trials are registered with ClinicalTrials.gov, numbers NCT00252733 for DIRECT-Prevent 1 and NCT00252720 for DIRECT-Protect 1. FINDINGS: 1421 participants (aged 18-50 years) were randomly assigned to candesartan (n=711) or to placebo (n=710) in DIRECT-Prevent 1, and 1905 (aged 18-55 years) to candesartan (n=951) or to placebo (n=954) in DIRECT-Protect 1. Incidence of retinopathy was seen in 178 (25%) participants in the candesartan group versus 217 (31%) in the placebo group. Progression of retinopathy occurred in 127 (13%) participants in the candesartan group versus 124 (13%) in the placebo group. Hazard ratio (HR for candesartan vs placebo) was 0.82 (95% CI 0.67-1.00, p=0.0508) for incidence of retinopathy and 1.02 (0.80-1.31, p=0.85) for progression of retinopathy. The post-hoc outcome of at least a three-step increase for incidence yielded an HR of 0.65 (0.48-0.87, p=0.0034), which was attenuated but still significant after adjustment for baseline characteristics (0.71, 0.53-0.95, p=0.046). Final ETDRS level was more likely to have improved with candesartan treatment in both DIRECT-Prevent 1 (odds 1.16, 95% CI 1.05-1.30, p=0.0048) and DIRECT-Protect 1 (1.12, 95% CI 1.01-1.25, p=0.0264). Adverse events did not differ between the treatment groups. INTERPRETATION: Although candesartan reduces the incidence of retinopathy, we did not see a beneficial effect on retinopathy progression.
|
|
