Researcher

Dr Rohitash Chandra

Keywords

Fields of Research (FoR)

Neural, Evolutionary and Fuzzy Computation, Earth Sciences, Environmental Technologies

Biography

Dr Rohitash Chandra is a Senior Lecturer in Data Science at the UNSW School of Mathematics and Statistics. Dr Chandra leads a program of research encircling methodologies and applications of artificial intelligence; particularly in areas of Bayesian deep learning, neuro-evolution,  climate extremes, geoscientific models, and mineral exploration.  Dr Chandra has developed novel methods for machine learning inspired by neural systems and learning...view more

Dr Rohitash Chandra is a Senior Lecturer in Data Science at the UNSW School of Mathematics and Statistics. Dr Chandra leads a program of research encircling methodologies and applications of artificial intelligence; particularly in areas of Bayesian deep learning, neuro-evolution,  climate extremes, geoscientific models, and mineral exploration.  Dr Chandra has developed novel methods for machine learning inspired by neural systems and learning behaviour that include transfer and multi-task learning, with the goal of modular deep learning. His current interest is uncertainty quantification and deep learning with applications to language models,  vaccine research, and COVID-19.

Dr Chandra has attracted multi-million dollar funding with a leading international interdisciplinary team. He is the Data Theme Lead of the Australian Research Council (ARC ITTC) Training Centre for Data Analytics in Minerals and Resources (2020-2025). Dr Chandra is an Associate Editor (Topical Editor) for Geoscientific Model Development, Neurocomputing (Elsevier), and  IEEE Transactions on Neural Networks and Learning Systems. Dr Chandra is a Senior Member of IEEE and UNSW Cultural Diversity Champion (2021-2023). Dr Chandra is part of NHMRC Medical Research Future Fund (2021-2022) for COVID-19 vaccine testing research led by CSIRO.  

Prior to joining UNSW, Dr Chandra held Sydney Research Fellowship at the University of Sydney (2017 - 2019). Prior to this, he has taken roles as Research Fellow in Machine Learning at Rolls Royce @Corp Lab, Nanyang Technological University, Singapore; Postdoctoral Research Fellow in Bioinformatics at Victoria University of Wellington (January to June 2012), and Lecturer in Computing Science at the University of the South Pacific (2013- 2015). Dr Chandra is originally from Fiji with a Girmit Indian heritage.

 

 

Available PhD project with scholarship

  • Bayesian deep learning for hydrological models (PhD) with Prof. Lucy Marshall. The project is funded by the Australian Research Council (ARC ITTC) Training Centre for Data Analytics in Minerals and Resources.
  • Requirements: First Class Honours or Masters by Research in Computational Statistics, Machine Learning, or Computer Science (with good grades). At least one paper in SJR Q1/Q2 ranked journal or equivalent conference proceeding. Available to Australian and international candidates. Apply: email rohitash.chandra at unsw.edu.au

Available Research Projects 

  1. Bayesian deep learning for protein function detection (PhD), Co-supervised by Prof. Alok Sharma
  2. Cyclone path and intensity prediction with deep insight based deep learning (Masters/Honours)
  3. Indoor path navigation for disabled persons in large buildings (Masters/Honours) 
  4. Detection of electric cable hazards from Cyclones using  drones and  remote sensing and deep learning (Masters/Honours) 
  5. Dynamic Earth models, landscape dynamics and basin evolution (PhD), Co-supervised by Prof. Dietmar Muller  More details
  6. Machine learning for reef modelling and Optimisation, Co-supervised by Prof. Jody Webster More information
  7. Deep learning for the reconstruction of 3D Ore-bodies, Co-supervised by  Dr Ehsan Farahbakhsh
  8. Memory in Recurrent Neural Networks and Neural Turing Machines, Honours/PhD
  9. Bayesian deep learning with incomplete information, Honours/PhD
  10. Variational Bayes for Spatio-temporal modelling (Honours/Masters/PhD) with Prof. Robert Kohn
  11. COVID-19 Modelling with Graph Neural Networks (Honours/Masters)
  12. Pruning Bayesian deep learning (Honours/Masters)
  13. Bayesian deep learning for language models  (Honours/Masters/PhD)
  14. Sentiment analysis with deep learning during natural disasters and extreme events (Honours/Masters/PhD)
  15. Human-robot language translation using deep learning  (Honours/Masters/PhD)
  16. Deep learning for monitoring abuse in social media  (Honours/PhD)
  17. Ensemble learning for class imbalanced problems (Honours/PhD)  with Dr Rodney Beard 
  18. Bayesian deep learning for hydrological models (Honours/PhD) with Prof. Lucy Marshall

 

Access research papers: https://github.com/rohitash-chandra/research

 

 


My Grants

  1. S. Vasan and R. Chandra et al. COVID-19 vaccine testing research. NHMRC Medical Research Future Fund  (MRFF) 2021-2022,  led by CSIRO.  
  2. S. Cripps, R. Chandra, et al., ARC training centre in data analytics for resources and environments (ARC ITTC DARE): UNSW Grant number: RG201207, 2020 - 2024. More information: https://darecentre.org.au/
  3. R. Chandra, Sydney Fellowship Awards, DVC Research, University of Sydney,  2017 -2019
  4. D. Muller, R. Chandra,  et al., Understanding the deep carbon cycle from icehouse to greenhouse climates, Sydney Research Excellence Initiative (SREI), DVC Research, University of Sydney, 2017 - 2018

My Qualifications

  1. PhD in Artificial Intelligence, Victoria University of Wellington (2012)
  2. MSc. in Artificial Intelligence, University of Fiji (2008)
  3. BSc. in Computer Science and Engineering Technology, University of the South Pacific (2006)

My Awards

  1. Doctoral Completion Award, Victoria University of Wellington (2012)
  2. Sydney Fellowship Award, University of Sydney (2017-2019)

My Research Activities

TECHNOLOGY DEVELOPMENT

 

Bayesian deep learning: Markov Chain Monte Carlo (MCMC) methods provide a probabilistic approach for the estimation of the free parameters in a wide range of models. Parallel tempering is an MCMC method that features parallelism with enhanced exploration capabilities. It features a number of replicas with slight variations in the acceptance criteria. We have developed Bayesian deep learning methods that feature parallel tempering and parallel computing. The challenge is in the inference for deep learning network architectures that features millions of parameters, such as convolutional neural networks and LSTM neural networks. Collaboration: Prof. Scott Sisson and Dr Pavel Krivitsky (UNSW)

Meta-learning and generative adversarial networks: The major challenge is to develop machine learning models given a low number of training examples. In this area, we use generative adversarial networks (GANs) with machine learning models to generative data in scenarios with space and limited data. The current focus is on pattern classification problems but the method can be used for spatiotemporal problems, and also augmented with Bayesian inference for robust uncertainty quantification. 

Surrogate-assisted and Bayesian optimisation: Surrogate-assisted optimization considers the estimation of an objective function for models given computational inefficiency or difficulty to obtain clear results. Surrogate-assistance inference addresses the inefficiency of parallel tempering for large-scale problems by combining parallel computing features with surrogate assisted estimation of likelihood function that describes the plausibility of a model parameter value, given specific observed data.  The challenge is to have a good estimation by the surrogates when the actual model features hundreds of free parameters. Collaboration: Prof. Dietmar Muller,  University of Sydney; Prof. Yew Soon Ong, Nanyang Technological University, Singapore.

Neuro-evolution and learning algorithms: Neural networks are loosely modelled after biological neural systems and have a wide range of data-driven applications that include time series prediction and pattern recognition. Opposed to gradient-based methods, neuro-evolution features evolutionary algorithms that provide a black-box approach to learning in neural networks. Hence, the learning algorithm is not constrained to the architecture of the network and does not face the limitations of gradient descent such as local minima and vanishing gradients. We have developed novel neural network learning algorithms using neuro-evolution with motivations from transfer learning, multi-task learning and reinforcement learning.  The challenge is in problems that have missing information, noise and inconsistencies in the organisation of data. Collaboration: Prof. Yew Soon Ong, Nanyang Technological University; Prof. Christian Omlin, University of Agder, Norway.

APPLICATIONS

Solid Earth evolution: Bayesian inference has been a popular methodology for the estimation and uncertainty quantification of parameters in geological and geophysical forward models. Badlands is a basin and landscape evolution forward model for simulating topography evolution at a large range of spatial and time scales. Our solid Earth evolution projects consider Bayesian inference for parameter estimation and uncertainty quantification for landscape dynamics model (Bayeslands). The challenge is in parameter estimation for computationally expensive models which are being addressed by high-performance computing and surrogate-assisted Bayesian inversion. Collaboration: Prof. Dietmar Muller and Dr Tristan Salles, University of Sydney.

Reef modelling: Geological reef models such as Py-Reef-Core provides insights into the flux of carbon by analysing carbonate platform growth and demise through time, and modelling their evolution using landscape dynamics and reef modelling. We estimate and provide uncertainty quantification of free model parameters using Bayesian inference with Py-Reef-Core. This can help us understand reef evolution on a geological timescale that can help in predicting the future evolution of coral reefs. The challenge here is in the estimation of the parameters which involves highly non-separable and constrained optimisation. Collaboration: A/Prof. Jody Webster and Dr Tristan Salles,  University of Sydney.

Mineral exploration: The extraction of geological lineaments from digital satellite data is a fundamental application in remote sensing. The location of geological lineaments such as faults and dykes are of interest in terms of mineralization. Although a wide range of applications utilized computer vision techniques, a standard workflow for the application of these techniques to mineral exploration is lacking. We use computer vision techniques for extracting geological lineaments using optical remote sensing data. Furthermore, in another research direction, we provide a synergy of geophysical forward models and Bayesian inference for 3D joint inversion for mineral prospecting and exploration. Collaboration: Prof. Dietmar Muller, Dr Ehsan Farahbakhsh 

Paleoclimate reconstruction: The reconstruction of paleoclimate precipitation can provide light to Earth’s climate history of millions of years in the past. Although global circulation models have been used with success for the reconstruction of precipitation in the Miocene period, their application to an era back in time is a major challenge due to limited data. We use an alternate approach that features machine learning methods to predict precipitation that defines paleoclimate that spans up to 400 million years in the past. The data features a range of geological indicators including sedimentary deposits (coal, evaporates, glacial deposits). The challenge has been in addressing missing values in the dataset and providing rigorous uncertainty quantification in order to develop paleo-maps of forests and vegetation. Collaboration: Prof. Dietmar Muller, Dr Nathaniel Butterworth, and Prof. Sally Cripps,  University of Sydney.

Cyclone modelling and prediction: The drastic effect of climate change is visible with extreme weather conditions such as tropical storms and cyclones. In this research, we use machine learning methods for forecasting cyclone formation for decades to come given drastic changes in the climate. We use global circulation models with machine learning methods to estimate cyclone categories decades ahead in the future. 

Language models: In this research, we use novel deep learning models to develop language models via social media to understand public behaviours in events such as COVID-19. We also develop language models to analyse translations of ancient Hindu texts that include the Bhagawad Gita and the Upanishads. 

 


My Research Supervision


Supervision keywords


Areas of supervision

Bayesian inference, deep learning,  machine learning, MCMC methods, Environmental informatics, Earth evolution models, reef evolution models, mineral exploration


Currently supervising

PhD research students 

  1. Hakiim Jamaluddin, ''Bayesian neural networks for unbalanced datasets``, School of Mathematics and Statistics, UNSW Sydney, from February 2021 (Joint Principal Supervisor with Prof. Scott Sisson)
  2. Subhash Chandra, " Machine learning and remote sensing  for mineral exploration", School of Minerals and Energy Resources Engineering, UNSW Sydney,  from February 2021 (Joint Principal Supervisor with Dr Stuart Clark)
  3. Ben Duthie, ''Bayesian deep learning for uncertain environments``, School of Mathematics and Statistics, UNSW Sydney, from September 2021 (Joint Principal Supervisor with Prof. Scott Sisson)
  4. Arpit Kapoor, ''Bayesian deep learning for hydrological models``, School of Civil and Environmental Engineering UNSW Sydney, from February 2022 (Joint Principal Supervisor with Prof. Lucy Marshall)
  5. Megan Nguyen, ''Leveraging expert knowledge and transfer learning with Bayesian deep learning ``, University of Sydney, from July 2022 (Co-Supervisor with Prof. Sally Cripps and Dr Milan Korbel)
  6. Ratneel Deo, ''Deep learning for understanding geo-coastal and reef development``, University of Sydney, from July 2022 (Co-Supervisor  with Prof. Jody Webster and Dr Tristan Salles)
  7. Robin Aldridge-Sutton, ''Deep learning for understanding bio-diversity and ecology``, University of Sydney, from July 2022 (Co-Supervisor  with Prof. Glenda Wardle and Dr Aaron Greenville)

Masters (Minor Thesis)  and Honours

  1. George Bai, "Bayesian neural ensemble learning with parallelized  and tempered Langevin MCMC",  Honours in Data Science, from 2021 (Principal Supervisor)
  2. Zhilin Wei, "Computer vision for aerial tracking of coastal plastic waste", Master of Statistics, UNSW Sydney, from May 2021 (Principal Supervisor, 2020) 
  3. Dizhou Feng, "Graph neural networks for spatiotemporal forecasting", Master of Statistics,  UNSW Sydney, from May 2021 (Principal Supervisor, 2020) 
  4. Mingyue Kang, "COVID-19 mutation over time", Master of Statistics,  UNSW Sydney, from September 2021 (Principal Supervisor, 2020) in collaboration with Prof. Seshadri Vasan (CSIRO)
  5. Jiaxin Yu, "COVID-19 diagnosis study with big data", Master of Statistics,  UNSW Sydney, from September 2021 (Principal Supervisor, 2020) in collaboration with Prof. Seshadri Vasan (CSIRO)

Postdoctoral Research Fellow

  • Dr Simon Luo, "Normalising flows based machine learning", University of Sydney, since June 2021 (Jointly supervised with Prof. Robert Kohn)

Recent completions

  1. Dr Ehsan Farahbakhsh, “Machine learning for mineral prospecting”, PhD, Amirkabir University of Technology, Tehran (External Supervisor, 2017 - 2020)
  2. Julian Rodriguez, "Machine learning for spatial-temporal mineral prospecting using plate tectonic models, MPhil, University of Sydney (External Supervisor with Prof. Dietmar Muller, 2019-2020)
  3. Yixuan He, "Bayesian neural learning for  financial prediction", Master of Financial Mathematics,  School of Mathematics and Statistics (Principal Supervisor, 2020) 
  4. Yueyang Zhang, "Gradient Boosting LSTM for reducing model uncertainty", Master of Statistics,  School of Mathematics and Statistics  (Principal Supervisor, 2021) 
  5. Shaodong Lin, "World economic outlook post-COVID-19 with deep learning", Master of Statistics,  School of Mathematics and Statistics  (Principal Supervisor, 2021) 

Research engineer 

  • Danial Azam. ARC Basin Genesis Hub, University of Sydney, Cosupervision with Prof. Dietmar Muller (Jan 2018 - December 2020)

Research interns 

      *Email Dr Chandra your CV if you wish to do an online research internship in machine learning. Open to local and international researchers/students.

2021

  1. Sweta Rathi, Indian Institute of Technology Guwahati, India (June - August 2021)
  2. Mukul Ranjan, Indian Institute of Technology Guwahati, India (June - August 2021)
  3. Amandeep Singh, Indian Institute of Technology Bombay, India (June - August 2021)
  4. Ritam Manabendra, Indian Institute of Technology Guwahati, India (June - August 2021)
  5. Anshul Negi, Indian Institute of Technology Roorkee, India (June - August 2021)
  6. Rishabh Sharma, Indian Institute of Technology Guwahati, India (June - August 2021)
  7. Sahil Bohra, Indian Institute of Technology Delhi, India (June - August 2021)
  8. Ayush Bhagat, Manipal Institute of Technology, India (April 2021 - June 2021)

2020

  1. Ritij Saini, Indian Institute of Technology Bombay, India (December 2020 - February 2021)
  2. Aswin Krishna, Indian Institute of Technology Guwahati, India (December 2020 - February 2021)
  3. Prabhat Singh, Indian Institute of Technology Guwahati, India (December 2020 - February 2021) - supervised jointly with Dr Anurag Sharma (University of South Pacific)
  4. Jiaxin Yu, School of Mathematics and Statistics, UNSW Sydney, (December 2020 - February 2021)
  5. Jiaxi Zhao, School of Mathematics and Statistics, UNSW Sydney, (December 2020 - February 2021)
  6. Animesh Renanse, Indian Institute of Technology Guwahati, India (May 2020 - August 2020)
  7. Shaurya Goyal, Indian Institute of Technoritamlogy Delhi, India (May 2020 - August 2020)
  8. Yash Sharma, Indian Institute of Technology Roorkee, India (May 2020 - August 2020)
  9. Ashish Gupta, Indian Institute of Technology Delhi, India (May 2020 - August 2020)
  10. Manavendrasinh Maharana, Manipal Institute of Technology, India (Jan 2020 - June 2020)
  11. Animesh Tiwari, Indian Institute of Technology Guwahati, India (May 2020 - August 2020)
  12. Eshwar Nukala, Indian Institute of Technology Guwahati, India (May 2020 - August 2020)
  13. Arya Arya Indian Institute of Technology Jammu, India ( August 2020 - December 2020) - supervised jointly with Prof Alok Sharma (RIKEN, Japan)
  14. Mahir Jain, Manipal Institute of Technology, India ( August 2020 - December 2020)
  15. Ayush Bhagat, Manipal Institute of Technology, India ( August 2020 - December 2020)
  16. Ayush Jain, Indian Institute of Technology Guwahati, India (May 2020 - August 2020)
  17. Divyanshu Singh, Indian Institute of Technology Guwahati, India ( August 2020 - December 2020)
  18. Kousik Rajesh, Indian Institute of Technology Guwahati, India (May 2020 - August 2020)

2019

  1. Aakarsh Yadav, Indian Institute of Technology, India (June 2019 - August 2019)
  2. Ashray Aman, Indian Institute of Technology Delhi, India (June 2019 - August 2019)
  3. Rishab Gupta, Indian Institute of Technology, India (June 2019 - August 2019)

 

2018

  1.  Konark Jain, Indian Institute of Technology, India (May 2018 - July 2018)
  2.  Arpit Kapoor, SRM Institute of Technology, India (June 2018 - August 2018)
  3.  Ratneel Deo, University of the South Pacific, Fiji (December 2017 - February 2018)
  4.  Wil Grebner, University of Sydney, Australia (February 2018 -  June 2018)

My Teaching

Master of Data Science: 

  1. ZZSC5836 - Data Mining and Machine Learning (Online), Hexamester 5 https://studyonline.unsw.edu.au/online-programs/master-data-science
  2. MATH5836 - Data Mining and its Business Applications, Trimester 3: https://www.maths.unsw.edu.au/courses/math5836-data-mining

 

 

View less

Location

School of Mathematics and Statistics
UNSW Sydney
NSW 2052
The Red Centre
Room 4110

Contact

0413071839
9385 7091