Browsing by Author "Ekanayake, IU"

Filter results by typing the first few letters
Now showing 1 - 5 of 5
  • Thumbnail Image
    item: Conference-Full-text
    Chronic Kidney Disease Prediction Using Machine Learning Methods
    (IEEE, 2020-07) Ekanayake, IU; Herath, D; Weeraddana, C; Edussooriya, CUS; Abeysooriya, RP
    Chronic Kidney Disease (CKD) or chronic renal disease has become a major issue with a steady growth rate. A person can only survive without kidneys for an average time of 18 days, which makes a huge demand for a kidney transplant and Dialysis. It is important to have effective methods for early prediction of CKD. Machine learning methods are effective in CKD prediction. This work proposes a workflow to predict CKD status based on clinical data, incorporating data prepossessing, a missing value handling method with collaborative filtering and attributes selection. Out of the 11 machine learning methods considered, the extra tree classifier and random forest classifier are shown to result in the highest accuracy and minimal bias to the attributes. The research also considers the practical aspects of data collection and highlights the importance of incorporating domain knowledge when using machine learning for CKD status prediction.
  • Thumbnail Image
    item:
    Explainable machine learning (XML) to predict external wind pressure of a low-rise building in urban-like settings
    (Elsevier, 2022) Meddage, DPP; Ekanayake, IU; Weerasuriya, AU; Lewangamage, CS; Tse, KT; Miyanawala, TP; Ramanayaka, CDE
    This study used explainable machine learning (XML), a new branch of Machine Learning (ML), to elucidate how ML models make predictions. Three tree-based regression models, Decision Tree (DT), Random Forest (RF), and Extreme Gradient Boost (XGB), were used to predict the normalized mean (Cp,mean), fluctuating (Cp,rms), minimum (Cp,min), and maximum (Cp,max) external wind pressure coefficients of a low-rise building with fixed dimensions in urban-like settings for several wind incidence angles. Two types of XML were used — first, an intrinsic explainable method, which relies on the DT structure to explain the inner workings of the model, and second, SHAP (SHapley Additive exPlanations), a post-hoc explanation technique used particularly for the structurally complex XGB. The intrinsic explainable method proved incapable of explaining the deep tree structure of the DT, but SHAP provided valuable insights by revealing various degrees of positive and negative contributions of certain geometric parameters, the wind incidence angle, and the density of buildings that surround a low-rise building. SHAP also illustrated the relationships between the above factors and wind pressure, and its explanations were in line with what is generally accepted in wind engineering, thus confirming the causality of the ML model's predictions.
  • Thumbnail Image
    item: Article-Full-text
    Modeling strength characteristics of basalt fiber reinforced concrete using multiple explainable machine learning with a graphical user interface
    (Nature Publishing Group, 2023) Kulasooriya, WKVJB; Ranasinghe, RSS; Perera, US; Thisovithan, P; Ekanayake, IU; Meddage, DPP
    This study investigated the importance of applying explainable artificial intelligence (XAI) on different machine learning (ML) models developed to predict the strength characteristics of basalt-fiber reinforced concrete (BFRC). Even though ML is widely adopted in strength prediction in concrete, the black-box nature of predictions hinders the interpretation of results. Among several attempts to overcome this limitation by using explainable AI, researchers have employed only a single explanation method. In this study, we used three tree-based ML models (Decision tree, Gradient Boosting tree, and Light Gradient Boosting Machine) to predict the mechanical strength characteristics (compressive strength, flexural strength, and tensile strength) of basal fiber reinforced concrete (BFRC). For the first time, we employed two explanation methods (Shapley additive explanations (SHAP) and local interpretable model-agnostic explanations (LIME)) to provide explanations for all models. These explainable methods reveal the underlying decision-making criteria of complex machine learning models, improving the end user's trust. The comparison highlights that tree-based models obtained good accuracy in predicting strength characteristics yet, their explanations were different either by the magnitude of feature importance or the order of importance. This disagreement pushes towards complicated decision-making based on ML predictions which further stresses (1) extending XAI-based research in concrete strength predictions, and (2) involving domain experts to evaluate XAI results. The study concludes with the development of a “user-friendly computer application” which enables quick st
  • Thumbnail Image
    item: Conference-Full-text
    Segmentation and significance of herniation measurement using lumbar intervertebral discs from the axial view
    (IEEE, 2022-07) Siriwardhana, Y; Karunarathna, D; Ekanayake, IU; Rathnayake, M; Adhikariwatte, V; Hemachandra, K
    According to statistics, more than 60% of people suffer lower back pain at a certain time in their lives. Disc hernias are the most common cause of lower back pain, and the lumbar spine is responsible for more than 95% of all herniated discs. Generally, radiologists study the MRI during the clinical phase to detect a disc hernia. There could be several cases to evaluate, leaving the doctors to cogitate and envisage. Medical image segmentation aids in the diagnosis of spinal pathology, studying the anatomical structures, surgical procedures, and the evaluation of various treatments. However, manual segmentation of medical images necessitates a significant amount of time, effort, and discipline on the part of domain experts. This research study describes a framework that automates the segmentation of lumbar intervertebral discs using MRI images. Through this system, we can detect minor changes at the pixel level that are impossible to identify with the naked eye. We used convolutional neural networks with the UNet architecture to achieve the semantic segmentation process. The segmentations were evaluated using the Jacquard index and the dice coefficient.
  • Thumbnail Image
    item: Conference-Full-text
    Tree-based regression models for predicting external wind pressure of a building with an unconventional configuration
    (IEEE, 2021-07) Meddage, DPP; Ekanayake, IU; Weerasuriya, AU; Lewangamage, CS; Adhikariwatte, W; Rathnayake, M; Hemachandra, K
    Traditional methods of pressure measurement of buildings are costly and time consuming. As an alternative to the traditional methods, this study developed a fast and computationally economical machine learning-based model to predict surface-averaged external pressure coefficients of a building with an unconventional configuration using three tree-based regressors: Adaboost, Extra Tree, and Random Forest. The accuracy and performance of the tree-based regressors were compared with a fourth-order polynomial function and a high-order non-linear regression proposed by an Artificial Neural Network (ANN). The comparison revealed random forest and extra tree models were simpler and more accurate than the polynomial functions and the ANN model. Alternatively, a machine learning interpretability method-Local Interpretable Model-agnostic Explanations (LIME) – was used to quantify the contribution of each parameter to the models' outcomes. LIME identified the most influential parameter, the variation in the influence of parameters with their values, and interactions of parameters. Moreover, LIME confirmed the tree-based regressors closely follow the flow physics in predicting external wind pressures rather than solely relied on training data.