Erandi Lakshika

Erandi Lakshika

Auto-generated profile, based on publicly available information

Erandi Lakshika

University of New South Wales, Canberra, ACT, Australia

Research Focus & Metrics

Academic Career Length

12Years (2015-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
9 August 2021 in IEEE Access
Verified
Journal Article
Open access
DSDilini Samarasinghe
Dilini Samarasinghe
MBMichael Barlow
Michael Barlow
ELErandi Lakshika
Erandi Lakshika
TLTimothy Lynar
Timothy Lynar
Nour MoustafaNour Moustafa
Nour Moustafa
BTBenjamin Peter Turnbull
Benjamin Peter Turnbull
Board games have often been recognised as a tool to model complex concepts in abstract environments for entertainment, education, and research in fields such as military and artificial intelligence. With more board games being designed and published, it is timely to draw attention towards board game design strategies and mechanics which capture the attributes that drive game play. The game design and the mechanics used define the structure, functionality, and play experience of these games. Towards this end, this paper presents a data driven review of board game mechanics and play-related attributes, their interactions and relationships. The analysis expects to draw insights into how board games can be utilised across diverse domains as a tool to understand and explore complex concepts through abstract models. The investigations focus on identifying the trends and patterns of board games being published and their individual mechanics over time. Moreover, the correlation between mechanics and play-related attributes such as game complexity, rating, and duration are explored. The interactions and similarities between individual mechanics based on co-occurrence, mutual information, and clustering based approaches are also illustrated. The results show that the level of complexity and engagement of a game is not a simple function of the set of mechanics used, but rather the interactions that exist between mechanics, and the nature of their specific implementation are the critical factors in determining play experience of a board game.

29 April 2016 in IEEE Transactions on Computational Intelligence and AI in Games
Verified
Journal Article
Closed access
ELErandi Lakshika
Erandi Lakshika
MBMichael Barlow
Michael Barlow
AEAdam Easton
Adam Easton
Modern video games come with highly realistic graphics enabling the players to interact with visually rich virtual worlds. Realistic (life-like) animation of nonplayer characters (NPCs) in such virtual worlds is particularly important to enhance the gaming experience. Multiagent systems are one effective approach to synthesize life-like behaviors and interactions by codifying simple rules into NPCs (each NPC as an autonomous agent). However, such behaviors generally come at the cost of increasing computational expense and complexity in terms of aspects such as number of rules and parameters. Therefore, the desire for high fidelity (highly realistic) behaviors is often in conflict with the drive for low complexity. Multiobjective evolutionary algorithms provide a sophisticated mechanism to optimize two or more conflicting objectives simultaneously. However, evolutionary computing techniques need an appropriate objective function to drive the exploration in the correct direction. Pairing of evolutionary techniques and multiagent systems is challenging in the classes of problems in which the fitness is evaluated based on human aesthetic judgment rather than on objective forms of measurements. In this study, we present a multiobjective evolutionary framework to evolve low complexity and high fidelity multiagent systems by utilizing a machine learning system trained by bootstrapping human aesthetic judgment. We have gathered empirical data in three problem areas—simulation of conversational group dynamics, sheepdog herding behaviors, and traffic dynamics, and show the effectiveness of our approach in deriving low complexity and high fidelity multiagent systems. Further, we have identified common properties of the Pareto-optimal frontiers in the three problem areas that can ultimately lead to an understanding of a relationship between simulation model complexity and behavior fidelity. This understanding will be useful in deciding which level of behavioral fidelity is required for the characters in video games based on the distance to the camera, importance to the scene, and available computational resources.

14 August 2023 in IEEE Access
Verified
Journal Article
Open access
SYShuo Yang
Shuo Yang
DSDilini Samarasinghe
Dilini Samarasinghe
AAAnupama Arukgoda
Anupama Arukgoda
SAShadi Abpeikar
Shadi Abpeikar
ELErandi Lakshika
Erandi Lakshika
MBMichael Barlow
Michael Barlow
Collective emergent behaviours are commonly seen in nature such as in flocks of birds and schools of fish. These behaviours are the results of years of evolution and have been studied in artificial agent systems in a wide range of application areas such as robotics, serious games, and crowd simulations. Automatic recognition of such collective behaviours is imperative in such application areas in order to measure and improve the effectiveness and efficiency of the artificial agent systems, especially when it involves machine learning approaches where human labelling is not feasible. While it is easy for the human eye to recognise collective behaviours, this is an extremely challenging task for a machine to automatically recognise them as such emergent behaviours cannot be captured by a simple mathematical equation. This paper investigates how emergent behaviours can be automatically recognised through capturing the behavioural aspects of the collective nature of the agents’ performance. We identify seven metrics such as grouping, order, and flock density that can capture diverse and distinct emergent characteristics of agent behaviours. Five machine learning models that use a combination of these metrics as features of a range of representative behaviours were trained to investigate the potential of automatic recognition of collective emergent behaviours. The evaluation results show that training the machine learning models with the proposed approach enables automatic recognition of a range of diverse emergent collective behaviours. Further, we conducted leave-one-behaviour-out experiments on the representative behaviours and the metrics used. The results confirmed that each behaviour and metric have a unique impact on accurate recognition of emergent behaviours in collective agent systems.

27 November 2020
Verified
Book Chapter
Closed access
PDF file
DWDulakshi Wannigamage
Dulakshi Wannigamage
MBMichael Barlow
Michael Barlow
ELErandi Lakshika
Erandi Lakshika
KKKathryn Kasmarik
Kathryn Kasmarik

26 September 2022 in IEEE Access
Verified
Journal Article
Open access
DSDilini Samarasinghe
Dilini Samarasinghe
MBMichael Barlow
Michael Barlow
ELErandi Lakshika
Erandi Lakshika
This paper presents a novel Flow-based reinforcement learning strategy to model agent systems that can adapt to complex and dynamic problem environments by incrementally mastering their skills. It is inspired by the psychological notion of Flow that describes the optimal mental state experienced by an individual when they are fully immersed in a task and find it intrinsically rewarding to engage with. The proposed model presents an algorithm to describe the Flow experience such that agents can be trained through finer distinctions to the challenges across training time to maintain them in the Flow zone. In contrast to the traditional and incremental learning approaches that suffer from limitations associated with overfitting, the Flow-based model drives agent behaviours not simply through external goals but also through intrinsic curiosity to improve their skills and thus the performance levels. Experimental evaluations are conducted across two simulation environments on a maze navigation task and a reward collection task with comparisons against a generic reinforcement learning model and an incremental reinforcement learning model. The results reveal that these two models are prone to overfit under different design decisions and loose the ability to perform in dynamic variations of the tasks in varying degrees. Conversely, the proposed Flow-based model is capable of achieving near optimal solutions with random environmental factors, appropriately utilising the previously learned knowledge to identify robust solutions to complex problems.

21 October 2022 in IEEE Transactions on Games
Verified
Journal Article
Closed access
CWChamila Wijayarathna
Chamila Wijayarathna
ELErandi Lakshika
Erandi Lakshika
In day-to-day life stress can arise due to various factors including work life demands, external situations and health issues. Stress becomes a concern when it affects a person's mental health, and sometimes it can even result in other chronic illnesses. Currently, stress detection studies in the literature are often limited to laboratory studies or to specific situations. However, daily stressors are on-going and therefore detection of stress in everyday life (outside the laboratory environments) is important to improve the wellbeing of individuals. Computer games is an entertainment media that can be found in almost every household in the modern society. Statistics show that people spend hours playing computer games daily. The amount of data that gameplay generates and interactivity they provide via various human computer interfaces have a lot of potential in identifying behaviour patterns of the players that could assist in the process of stress detection. As such, this survey attempts to identify the extent to which computer games can be used as a medium for stress detection. Towards this end, this survey reviews the existing stress detection studies, both laboratory techniques, as well as the techniques that can be used in a home-based environment. Finally, it summarises the stress detection techniques that can be used within games in order to make it an everyday technology that can be used to detect and monitor stress. In addition, it is expected that development of such a technology will be useful in providing objective data to the health care professionals for intervention and management. Such a technology is even more required in the current unprecedented situation the world has faced due to the COVID-19 pandemic as it can be developed as a technology to manage mental health issues people are facing due to home isolation.