| 1. |
- Fanaee Tork, Hadi, 1983-, et al.
(författare)
-
Iterative Multi-mode Discretization : Applications to Co-clustering
- 2020
-
Konferensbidrag (refereegranskat)abstract
- We introduce a new concept called “Iterative Multi-Mode Discretization (IMMD)” which is a new type of efficient data sparsification that can scale up many tasks in data mining. In this paper we demonstrate the application of IMMD in co-clustering, i.e. simultaneous clustering of the rows and columns in a matrix. We propose IMMD-CC, a novel co-clustering algorithm, which is developed based on IMMD. IMMD-CC has attractive properties. First, its time complexity is linear, so it can be used in large-scale problems. In addition, IMMD-CC is able to estimate the number of co-clusters automatically, and more accurate than state-of-the-art methods. We demonstrate the performance of IMMD-CC in comparison to several state-of-the-art methods on 100 data sets from a benchmark cohort, as well as 35 real-world datasets. The results show the promising potential of the proposed method.
|
|
| 2. |
- Leder, Lena, et al.
(författare)
-
Effects of a healthy Nordic diet on gene expression changes in peripheral blood mononuclear cells in response to an oral glucose tolerance test in subjects with metabolic syndrome : A SYSDIET sub-study
- 2016
-
Ingår i: Genes & Nutrition. - : Springer Science and Business Media LLC. - 1555-8932 .- 1865-3499. ; 11:1
-
Tidskriftsartikel (refereegranskat)abstract
- Background: Diet has a great impact on the risk of developing features of metabolic syndrome (MetS), type 2 diabetes mellitus (T2DM), and cardiovascular diseases (CVD). We evaluated whether a long-term healthy Nordic diet (ND) can modify the expression of inflammation and lipid metabolism-related genes in peripheral blood mononuclear cells (PBMCs) during a 2-h oral glucose tolerance test (OGTT) in individuals with MetS. Methods: A Nordic multicenter randomized dietary study included subjects (n = 213) with MetS, randomized to a ND group or a control diet (CD) group applying an isocaloric study protocol. In this sub-study, we included subjects (n = 89) from three Nordic centers: Kuopio (n =26), Lund (n = 30), and Oulu (n = 33) with a maximum weight change of ±4 kg, high-sensitivity C-reactive protein concentration ≤10 mg L-1, and baseline body mass index -2. PBMCs were isolated, and the mRNA gene expression analysis was measured by quantitative real-time polymerase chain reaction (qPCR). We analyzed the mRNA expression changes of 44 genes before and after a 2hOGTT at the beginning and the end of the intervention. Results: The healthy ND significantly down-regulated the expression of toll-like receptor 4 (TLR4), interleukin 18 (IL18), and thrombospondin receptor (CD36) mRNA transcripts and significantly up-regulated the expression of peroxisome proliferator-activated receptor delta (PPARD) mRNA transcript after the 2hOGTT compared to the CD. Conclusions: A healthy ND is able to modify the gene expression in PBMCs after a 2hOGTT. However, more studies are needed to clarify the biological and clinical relevance of these findings.
|
|
| 3. |
- Zhang, Chi, et al.
(författare)
-
Feature extraction from unequal length heterogeneous EHR time series via dynamic time warping and tensor decomposition
- 2021
-
Ingår i: Data mining and knowledge discovery. - New York, NY : Springer. - 1384-5810 .- 1573-756X. ; 35, s. 1760-1784
-
Tidskriftsartikel (refereegranskat)abstract
- Electronic Health Records (EHR) data is routinely generated patient data that can provide useful information for analytical tasks such as disease detection and clinical event prediction. However, temporal EHR data such as physiological vital signs and lab test results are particularly challenging. Temporal EHR features typically have different sampling frequencies; such examples include heart rate (measured almost continuously) and blood test results (a few times during a patient’s entire stay). Different patients also have different length of stays. Existing approaches for temporal EHR sequence extraction either ignore the temporal pattern within features, or use a predefined window to select a section of the sequences without taking into account all the information. We propose a novel approach to tackle the issue of irregularly sampled, unequal length EHR time series using dynamic time warping and tensor decomposition. We use DTW to learn the pairwise distances for each temporal feature among the patient cohort and stack the distance matrices into a tensor. We then decompose the tensor to learn the latent structure, which is consequently used for patient representation. Finally, we use the patient representation for in-hospital mortality prediction. We illustrate our method on two cohorts from the MIMIC-III database: the sepsis and the acute kidney failure cohorts. We show that our method produces outstanding classification performance in terms of AUROC, AUPRC and accuracy compared with the baseline methods: LSTM and DTW-KNN. In the end we provide a detailed analysis on the feature importance for the interpretability of our method. © 2021, The Author(s), under exclusive licence to Springer Science+Business Media LLC, part of Springer Nature.
|
|
