CHAMIN NALINDA LOKUGAM HEWAGE
I am a Computer Science PhD student at University College Dublin, Ireland under Dr. Michela Bertolotto and Dr. Nhien-An Le-Khac. I liases with Dr. Anh Vu in my research work and he too is a great asset. My main research focus involves management of LiDAR data in distributed cloud environment (More info will be available soon at my Medium blog).
I received my BSc.(Hons.)(First Class) in Computer Science from University of Colombo, School of Computing, Sri Lanka. Prior to join as a PhD student, I worked at Sysco LABS, Sri Lanka as a Full-Stack Software Engineer for one year and as a Research Engieer at Nanyang Technical University, Singapore for two years.
Major: LiDAR, Spatial and Temporal Databases, Distributed Cloud Storage (Hadoop, HBase)
RESEARCH
Update in Progress!
PhD Work (update in progress)CISCO Project (update in progress)
Modelling and Optimizing Homecare and Caregiving Services for the Elderly
Nanyang Technological University (NTU), Singapore
Supervisor: Associate Professor Ng Wee Keong (PhD. University of Michigan)
Responsibilities:
- Develop algorithms and data structures to effectively analyze data (historical, EDA and real-time) that obtained from the health-care domain.
- Develop recommendation system (Constraint based)
- Develop a robust, low cost, long range “Home-Care-Monitoring system” (LoRaWAN).
Contribution:
Markov Model for anomaly detection, Naïve Bayes for patient profiling, ‘Constraint Based Recommendation System’ development, Literature review, architecture design, device design, implementation of algorithms and data structures considering Machine Learning, development of scalable web application (micro services).
Domain Specific Language for Specifying Operations of a Central Counterparty
University of Colombo, Sri Lanka (Final Year Undergraduate Research)
Supervisor: Dr. Chamath Keppitiyagama(PhD. University of British Columbia)
Research Summary:
- The aim of this research was to extend the applicability of Haskell Contract Combinator Library (a functional programming library to define financial contracts), to the operations of a financial institute called Central Counterparty (CCP).
- Contribution: Literature review, define cash flow semantics of CCP operations using Haskell library, conceptualization of “seed contract”, conceptualization of “Margin Calculation Engine for CCP”.
- Publication: Research paper titled “Domain Specific Language for Specifying Operations of a Central Counterparty”(https://goo.gl/eSrFsi), accepted and published at 17th International Conference of Advances in ICT for Emerging Regions (ICTer2017)(IEEE).
PROJECTS
"fast.ai" Deep Learning course by Jeramy Howard
My goal is to get hands on experiences in deep learning. This course has designed targetting developers and coders. I purchased Amazon p2.xlarge instance to train and develop the models. The pet projects that taught in this course help me to strengthen my understanding in deep learning, specially in Computer Vision related.
[Python Data Science, AWS, Keras, Caffe, Neural Networks]
"Kaggle" Competitions
I learn from peers who have scored high scores in competitions. While learning, I try to fill in the gaps in my knowledge. I consider "Kaggle" as an excellent platform to get familiar with practical aspectis of Machine Learning, Data Visualization and other data science related concepts.
Codes are hosted @ GitHub.
Software Engineer @ Sysco LABS, Colombo, Sri Lanka.
Sysco Pulse Application
Mainly contributed in front-end, back-end development, “Salesforce” API integration, Python client authentication using OAuth, network optimization through secure web sockets (WSS), product feature implementation to execute Python and Shell commands remotely through HTTS and WSS.
[Python, Node JS, Web Sockets, Linux, Couch DB, MySQL, SSH, Shell scripting, Angular JS, React JS, OAuth, OpenID Connect, AWS, Salesforce]
Software Engineer Intern @ WSO2 Inc., Colombo, Sri Lanka.
Credit Card Fraud Detection Toolbox for WSO2 Complex Event Processor (WSO2 CEP)
Supervisor: Dr. Srinath Pereran (PhD. University of Indiana, Bloomington)(Director of Research @ WSO2)
WSO2 Complex-Event-Processor (CEP) which is an Apache Storm based analytic engine. In this task, I implemented Markov Model (MM), which was an 18x18 probability matrix. I used statistical matrices (Miss Probability, Miss Rate Probability, and Entropy Reduction) in evaluating if the transactions are fraudulent. The developed MM continuously learns through the event streams that comes to CEP. I further used in-memory tables to improve the performance.
[Java Web Services, Maven, Middleware, SOA, Apache Storm, Complex-Event-Processing, Distributed Stream Processing, Real-time Analytics]
Service Statistic Dashboard for WSO2 Identity Server (WSO2 IS)
Supervisor: Mr. Prabath Siriwardena (Director Security @ WSO2)
Implemented an Axis2 Handler and a Tomcat Valve to capture IS services data such as the user, the application, the authentication protocol used (SAML, XACML, OAuth, OpenID Connect), application login time, logout time, success/fail attempts. These captured data were then published to WSO2 Business Activity Monitor (BAM) Cassandra column family, using a Log Event Appender. These unstructured data that resides in Cassandra columns were published to MySQL database using a Hive query.
[Java Web Services, Maven, SOA, Cassandra, Hive, MySQL, Identity and Security Management]
Undergraduage Projects
"Sidhadhiya" web and mobile application (Grade A)
Supervisor: Dr. D.D. Karunaratne (PhD. University of Cardiff)
I did front-end development using Twitter Bootstrap CSS framework to make the website cross-browser and cross device compatible. I used PHP5 for the server side development and use MySQLi as the database infrastructure. Also I integrated the developed Android application with the Google Cloud Messaging (GCM). This is a group project and I was the group leader.
Augmented Reality Android application (Grade A)
Supervisor: Dr. G.D.S.P. Wimalaratne (PhD. Salford University)
Android phone user will see an augmented environment on phone screen when user direct the phone at the image target. Sound level will increase/decrease depending on the proximity to the image target. User can interact with the augmented environment by dragging, rotating, scaling the objects in the environment (https://goo.gl/sUqkUt).
[Unity 3D, Vuforia, C#]
PUBLICATIONS
Conference and Workshop Papers
[1] Chamin Nalinda, Chamath Keppitiyagama. Domain Specific Language for Speficiying Operations of a Central Counterparty, ICTER 2017, September 2017.
Technical Reports
[2] Chamin Nalinda. Big Data Social Network Analysis, December 2014.
AWARDS
[1] AI Scholar at Pi School in Rome, Italy. Won a full grant scholarship to participate a two month study program in Artificial Intelligence at Pi School in Rome, Italy.
[2] University GOLD Medal on Best Performance in Industrial Placement, awarded at University convocation 15th August 2016.
VOLUNTEER ACTIVITIES
Google Code-in mentor for Sustainable Computing Research Group (SCoRe) 2016
SL2College Administrator for "Mentor Team"2014-2016