Rajith Vidanaarachchi

Rajith Vidanaarachchi

Post-Doc Researcher and Educator at the University of Melbourne

Talks and presentations

Fairness and Ethics in AI: Ethical Responsibility of Computer Scientists

January 01, 2024

Invited Talk, AI Horizons: Exploring Real-world AI Applications, University of Jaffna, University of Melbourne, and Google Research

Presented on fairness and ethics in AI, discussing the ethical responsibility of computer scientists when developing and deploying real-world AI applications. This talk was part of the AI Horizons series, a collaboration between the University of Jaffna, the University of Melbourne, and Google Research.

An Agent-Based Model based on the Theory of Planned Behaviour with Long-Short Term Memory Modules — Case Study on Food Delivery and Its Implications

September 01, 2023

Conference Presentation, Social Simulation Conference (SSC) 2023, Glasgow, UK

Presented research on integrating Long-Short Term Memory (LSTM) neural network modules with an agent-based model grounded in the Theory of Planned Behaviour. The case study examined food delivery systems and their broader implications for urban planning and policy.

Agent Based Modelling for Understanding Social Systems

July 01, 2023

Invited Workshop, ARC Centre of Excellence for Children and Families over the Life Course - Seminar Series, Melbourne, Australia

Invited workshop on agent-based modelling approaches for understanding complex social systems, covering methodology, tools, and applications in social science research.

Fairness and Ethics in Generative AI

June 01, 2023

Guest Lecture, The University of Melbourne, Melbourne, Australia

Delivered a guest lecture on fairness and ethics in generative AI at the University of Melbourne, covering the emerging ethical challenges posed by large language models and generative systems, including issues of bias, representation, and responsible deployment.

Fairness in AI: Understanding Biases

March 01, 2023

Visiting Lecture, University of Peradeniya, Peradeniya, Sri Lanka

Delivered a visiting lecture on fairness in AI at the University of Peradeniya, discussing how biases arise in AI systems, their societal implications, and approaches to building fairer machine learning models.

DALLE-URBAN: Capturing the Urban Design Expertise of Large Text to Image Transformers

December 01, 2022

Conference Presentation, International Conference on Digital Image Computing: Techniques and Applications (DICTA 2022), Sydney, Australia

Presented our investigation into the capabilities and biases of text-to-image methods as they apply to the built environment, evaluating how large generative models capture urban design knowledge.

Fairness in AI for Health

July 01, 2022

Conference Presentation, Moratuwa Engineering Research Conference (MerCon) 2022, Virtual

Presented on fairness in AI applied to healthcare contexts, discussing how algorithmic biases can disproportionately affect health outcomes and exploring approaches to ensure equitable AI systems in medical and public health applications.

AgentsX.jl — An Extended Julia Framework for Exploring Urban and Social Systems

May 10, 2022

Workshop Presentation, International Workshop on Agent-Based Modelling of Urban Systems (ABMUS) at AAMAS 2022, Online

Presented the key design ideas for AgentsX.jl, a proposed extension and interface to the Agents.jl framework for agent-based modelling of urban and social systems simulation, formalising ABM design through a “Code as the Model” approach.

Towards Modelling Energy Demand of Vehicles in Cities: An Agent-Based Method

April 02, 2022

Conference Presentation, 30th Annual Geographical Information Science Research UK (GISRUK) Conference, Liverpool, UK

Presented our agent-based model of vehicle activity for quantifying electric energy consumption through the simulation of a heterogeneous fleet of electric vehicles within an urban street network.

Imagery, Privacy and Ethics: An Overview of Partially Occluded Facial Biometric Analysis in the Era of Face Masks

December 01, 2021

Conference Presentation, International Conference on Electrical, Computer and Energy Technologies (ICECET 2021), Prague, Czech Republic

Presented the first review of masked facial recognition tasks and techniques robust to masked facial images, providing a central reference for automated public health and COVID-safe identification protocols while exploring the ethical aspects of these techniques.

Real Time Parameter Changes in Agent-Based Models with Mesa

September 01, 2021

Conference Presentation, Conference of the European Social Simulation Association (ESSA 2021), Online

Presented our work on enabling real-time parameter modification within Mesa-based agent-based model simulations, enhancing interactive exploration of complex system dynamics.

From The International Space Station To Tropical Rainforests And Polar Ice Caps: Microbial Communities Foretell The Effects Of Climate Change

August 01, 2021

Conference Presentation, 10th International Conference on Information and Automation for Sustainability (ICIAfS 2021), Sri Lanka

Presented our study examining how microbial communities across diverse environments — from the International Space Station to tropical rainforests and polar ice caps — can foretell the effects of climate change.

DynMIN: Temporally Dynamic Microbial Interaction Network Inference

June 01, 2020

Conference Presentation, RECOMB 2020, Virtual

Presented DynMIN, a method for inferring temporally dynamic microbial interaction networks from time-series metagenomic data. The talk covered the computational approach to capturing how microbial community interactions evolve over time, providing insights into complex biological system dynamics.

Exploring the Dynamics of Microbial Interactions

February 01, 2020

Conference Presentation, Lorne Genome Conference 2020, Lorne, Australia

Presented research on exploring the dynamics of microbial interactions using computational approaches, discussing methods for understanding how microbial communities change and interact over time in diverse environments.

IMPARO: Inferring Microbial Interactions through Parameter Optimisation

December 19, 2019

Conference Presentation, International Conference on Bioinformatics (InCoB) 2019, Jakarta, Indonesia

Presented IMPARO, a computational method for inferring microbial interactions through parameter optimisation. The talk covered the approach to modelling microbial community dynamics by optimising interaction parameters from time-series abundance data.

Exploring Computational Inference of Microbial Interactions and their Dynamics

December 18, 2019

Conference Presentation, 14th Conference on Industrial and Information Systems (ICIIS 2019), Kandy, Sri Lanka

Presented our work exploring current approaches in inferring microbial interactions and demonstrating their dynamic nature, with a proposed visualisation approach for observing microbial interaction network dynamics.

Data-driven Insights for Complex Biological Systems

November 01, 2019

Conference Presentation, Institution of Engineers Sri Lanka (IESL) 2019, Colombo, Sri Lanka

Presented on data-driven approaches for gaining insights into complex biological systems, discussing computational methods for analysing and modelling the behaviour of biological communities using large-scale data.