| 1. |
- Nguyen, Quang Hung, et al.
(författare)
-
Influence of Data Splitting on Performance of Machine Learning Models in Prediction of Shear Strength of Soil
- 2021
-
Ingår i: Mathematical problems in engineering (Print). - UK : Hindawi Publishing Corporation. - 1024-123X .- 1563-5147. ; 2021, s. 1-15
-
Tidskriftsartikel (refereegranskat)abstract
- The main objective of this study is to evaluate and compare the performance of different machine learning (ML) algorithms, namely, Artificial Neural Network (ANN), Extreme Learning Machine (ELM), and Boosting Trees (Boosted) algorithms, considering the influence of various training to testing ratios in predicting the soil shear strength, one of the most critical geotechnical engineering properties in civil engineering design and construction. For this aim, a database of 538 soil samples collected from the Long Phu 1 power plant project, Vietnam, was utilized to generate the datasets for the modeling process. Different ratios (i.e., 10/90, 20/80, 30/70, 40/60, 50/50, 60/40, 70/30, 80/20, and 90/10) were used to divide the datasets into the training and testing datasets for the performance assessment of models. Popular statistical indicators, such as Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), and Correlation Coefficient (R), were employed to evaluate the predictive capability of the models under different training and testing ratios. Besides, Monte Carlo simulation was simultaneously carried out to evaluate the performance of the proposed models, taking into account the random sampling effect. The results showed that although all three ML models performed well, the ANN was the most accurate and statistically stable model after 1000 Monte Carlo simulations (Mean R = 0.9348) compared with other models such as Boosted (Mean R = 0.9192) and ELM (Mean R = 0.8703). Investigation on the performance of the models showed that the predictive capability of the ML models was greatly affected by the training/testing ratios, where the 70/30 one presented the best performance of the models. Concisely, the results presented herein showed an effective manner in selecting the appropriate ratios of datasets and the best ML model to predict the soil shear strength accurately, which would be helpful in the design and engineering phases of construction projects.
|
|
| 2. |
- Wynberg, Elke, et al.
(författare)
-
Variability in white blood cell count during uncomplicated malaria and implications for parasite density estimation : a WorldWide Antimalarial Resistance Network individual patient data meta-analysis
- 2023
-
Ingår i: Malaria Journal. - : Springer Nature. - 1475-2875. ; 22
-
Tidskriftsartikel (refereegranskat)abstract
- Background: The World Health Organization (WHO) recommends that when peripheral malarial parasitaemia is quantified by thick film microscopy, an actual white blood cell (WBC) count from a concurrently collected blood sample is used in calculations. However, in resource-limited settings an assumed WBC count is often used instead. The aim of this study was to describe the variability in WBC count during acute uncomplicated malaria, and estimate the impact of using an assumed value of WBC on estimates of parasite density and clearance.Methods: Uncomplicated malaria drug efficacy studies that measured WBC count were selected from the WorldWide Antimalarial Resistance Network data repository for an individual patient data meta-analysis of WBC counts. Regression models with random intercepts for study-site were used to assess WBC count variability at presentation and during follow-up. Inflation factors for parasitaemia density, and clearance estimates were calculated for methods using assumed WBC counts (8000 cells/mu L and age-stratified values) using estimates derived from the measured WBC value as reference.Results: Eighty-four studies enrolling 27,656 patients with clinically uncomplicated malaria were included. Geometric mean WBC counts (x 1000 cells/mu L) in age groups < 1, 1-4, 5-14 and >= 15 years were 10.5, 8.3, 7.1, 5.7 and 7.5, 7.0, 6.5, 6.0 for individuals with falciparum (n = 24,978) and vivax (n = 2678) malaria, respectively. At presentation, higher WBC counts were seen among patients with higher parasitaemia, severe anaemia and, for individuals with vivax malaria, in regions with shorter regional relapse periodicity. Among falciparum malaria patients, using an assumed WBC count of 8000 cells/mu L resulted in parasite density underestimation by a median (IQR) of 26% (4-41%) in infants < 1 year old but an overestimation by 50% (16-91%) in adults aged = 15 years. Use of age-stratified assumed WBC values removed systematic bias but did not improve precision of parasitaemia estimation. Imprecision of parasite clearance estimates was only affected by the within-patient WBC variability over time, and remained < 10% for 79% of patients.Conclusions: Using an assumed WBC value for parasite density estimation from a thick smear may lead to underdiagnosis of hyperparasitaemia and could adversely affect clinical management; but does not result in clinically consequential inaccuracies in the estimation of the prevalence of prolonged parasite clearance and artemisinin resistance.
|
|
