Research Publications Authored by SLIIT Staff
Permanent URI for this communityhttps://rda.sliit.lk/handle/123456789/4195
This collection includes all SLIIT staff publications presented at external conferences and published in external journals. The materials are organized by faculty to facilitate easy retrieval.
Browse
3 results
Filters
Advanced Search
Filter by
Settings
Search Results
Publication Embargo LISA : Enhance the explainability of medical images unifying current XAI techniques(IEEE, 2022-07-18) Abeyagunasekera, S. H. P; Perera, Y; Chamara, K; Kaushalya, U; Sumathipala, PThis work proposed a unified approach to increase the explainability of the predictions made by Convolution Neural Networks (CNNs) on medical images using currently available Explainable Artificial Intelligent (XAI) techniques. This method in-cooperates multiple techniques such as LISA aka Local Interpretable Model Agnostic Explanations (LIME), integrated gradients, Anchors and Shapley Additive Explanations (SHAP) which is Shapley values-based approach to provide explanations for the predictions provided by Blackbox models. This unified method increases the confidence in the black-box model’s decision to be employed in crucial applications under the supervision of human specialists. In this work, a Chest X-ray (CXR) classification model for identifying Covid-19 patients is trained using transfer learning to illustrate the applicability of XAI techniques and the unified method (LISA) to explain model predictions. To derive predictions, an image-net based Inception V2 model is utilized as the transfer learning model.Publication Embargo A reinforcement learning approach to enhance the trust level of MANETs(IEEE, 2018-10-02) Jinarajadasa, G; Rupasinghe, L; Murray, IA Mobile ad-hoc network (MANET) consists of many freely interconnected and autonomous nodes which are often composed of mobile devices. MANETs are decentralized and self-organized wireless communication systems which are able to arrange themselves in various ways and have no fixed infrastructure. Since MANETs are mobile, the network topology changes rapidly and unpredictably. Because of this nature of the mobility of the nodes in MANETs, the main problems that occur are, unreliable communications and weak security where the data can be compromised or easily misused. Therefore, we propose a trust enhancement approach to a MANET, which is based on RLTM (Reinforcement Learning Trust Manager), a set of algorithms that considers Ad-hoc On-demand Distance Vector (AODV) protocol as the specific protocol, via Reinforcement Learning (RL) and Deep Learning concepts. The proposed system consists of an RL agent that, learns to detect and give predictions on trustworthy nodes, reputed nodes, and malicious nodes and to classify them. The identified parameters from AODV simulation using Network Simulator-3(NS-3) were given to the designed RNN (Recurrent Neural Network) model and results were evaluated.Publication Embargo Smartphone-based approach to enhance mindfulness among undergraduates with stress(IEEE, 2019-12-18) Edirisooriya, Y. L; Rathnayake, N. A; Ariyasena, T. D; Thelijjagoda, S; Jayawickrama, N. D; Chamindi, D. IOver the last decade, creating awareness among people has become a common part of the curriculum around the world. To date, researchers have predominately used outcome-based trial designs to understand the efficacy of mindfulness for improving wellness in people. Less research has been directed towards understanding how undergraduates perceive mindfulness experiences. Findings suggest that mindfulness enhances human wellbeing and helps people to develop a greater awareness of their body, mind, and emotions. The main objective of the research was to develop an application, an undergraduate could use to release their stress in tough situations using mindfulness practices. The conducted research project includes VR technologies, Reinforcement learning, Wearable Arduino components together with mindfulness-based stress reduction methods to support the features of the designed application. It may, therefore, provide a key to understanding one of the mechanisms that underlie the observed benefits of mindfulness approaches for mental health and wellbeing. Drawing on a sample of 25 questionnaire-based responses from the users, the generated results indicate the developed algorithm has an accuracy of 96.153%. The conducted research Evidently shows the use of reinforcements learning is a better way of predicting user interests. Implications of these findings are discussed further throughout the study. Moreover, mindfulness stress reduction practices for university undergraduates offer systematic relaxation methods and concrete ways to get rid of recurrent worries, reducing stress and its physical effects.