Dr Sharu Jose PhD

Dr Sharu Jose

School of Computer Science
Assistant Professor

Contact details

Email
s.t.jose@bham.ac.uk
Twitter
@sharu_theresa
View my research portal
Address
School of Computer Science
University of Birmingham
Edgbaston
Birmingham
B15 2TT
UK

Sharu Jose is an Assistant Professor in the Department of Computer Science. Her research interests lie broadly at the intersection of information theory, learning theory and machine learning. In particular, she is interested in information-theoretic generalization analysis of classical and quantum machine learning models, hybrid quantum-classical variational algorithms (convergence/generalization) and quantum error mitigation.

Qualifications

  • Ph.D. in Information Theory, 2018 (Indian Institute of Technology Bombay, India)
  • M.Tech in Systems and Control Engineering, 2013
    (Indian Institute of Technology Roorkee, India)
  • B.Tech in Electrical and Electronics Engineering, 2011
    (University of Kerala, India)

Biography

Dr. Sharu Theresa Jose received her Ph.D. from Indian Institute of Technology Bombay (IITB), India in August of 2018. Her thesis was on finite blocklength information theory with her main contribution being  a novel linear programming-based framework to develop converses for coding problems in information theory. During the course of her Ph.D., she published 3 articles in the IEEE Transactions on Information Theory, and was awarded the Best Ph.D. Thesis Award from IITB.

From October 2019 to July 2022, she was a postdoctoral researcher with Prof. Osvaldo Simeone of King's College London working on problems at the intersection of information theory and machine learning.

Postgraduate supervision

Quantum Machine Learning - Theoretical and Algorithmic Foundations

Research

Her broad interests lie at the intersection of machine learning (classical and quantum), statistical learning theory and information theory. She aims to use information-theory to advance theoretical understanding of learning problems as well as algorithm development. Some of her recent works include:

1. Information-theoretic generalization analysis of multi-task learning problems that include transfer learning and meta-learning

2. Quantum machine learning - generalization analysis, convergence analysis of variational quantum algorithms in NISQ devices, quantum error mitigation

Publications

Recent publications

Article

Jose, ST & Simeone, O 2022, 'An information-theoretic analysis of the cost of decentralization for learning and inference under privacy constraints', Entropy, vol. 24, no. 4, 485. https://doi.org/10.3390/e24040485

Jose, ST, Simeone, O & Durisi, G 2022, 'Transfer meta-learning: information- theoretic bounds and information meta-risk minimization', IEEE Transactions on Information Theory, vol. 68, no. 1, pp. 474-501. https://doi.org/10.1109/TIT.2021.3119605

Jose, ST & Simeone, O 2021, 'Free energy minimization: a unified framework for modeling, inference, learning, and optimization [lecture notes]', IEEE Signal Processing Magazine, vol. 38, no. 2, pp. 120-125. https://doi.org/10.1109/MSP.2020.3041414

Jose, ST & Simeone, O 2021, 'Information-theoretic generalization bounds for meta-learning and applications', Entropy, vol. 23, no. 1, 126. https://doi.org/10.3390/e23010126

Jose, ST & Kulkarni, AA 2020, 'Shannon Meets von Neumann: A Minimax Theorem for Channel Coding in the Presence of a Jammer', IEEE Transactions on Information Theory. https://doi.org/10.1109/TIT.2020.2971682

Jose, ST & Kulkarni, AA 2019, 'Improved Finite Blocklength Converses for Slepian–Wolf Coding via Linear Programming', IEEE Transactions on Information Theory. https://doi.org/10.1109/TIT.2018.2873623

Conference contribution

Jose, S, Park, S & Simeone, O 2022, Information-theoretic analysis of epistemic uncertainty in Bayesian meta-learning. in G Camps-Valls, FJR Ruiz & I Valera (eds), International Conference on Artificial Intelligence and Statistics, 28-30 March 2022, A Virtual Conference. Proceedings of Machine Learning Research, vol. 151, Proceedings of Machine Learning Research, pp. 9758-9775, The 25th International Conference on Artificial Intelligence and Statistics, 28/03/22. <https://proceedings.mlr.press/v151/theresa-jose22a.html>

Jose, S & Simeone, O 2021, A unified PAC-Bayesian framework for machine unlearning via information risk minimization. in 2021 IEEE 31st International Workshop on Machine Learning for Signal Processing (MLSP)., 9596170, Machine learning for signal processing , IEEE. https://doi.org/10.1109/MLSP52302.2021.9596170

Jose, S & Simeone, O 2021, An information-theoretic analysis of the impact of task similarity on meta-learning. in 2021 IEEE International Symposium on Information Theory (ISIT)., 9517767, IEEE International Symposium on Information Theory proceedings, IEEE, pp. 1534-1539, 2021 IEEE International Symposium on Information Theory, Melbourne, Victoria, Australia, 12/07/21. https://doi.org/10.1109/ISIT45174.2021.9517767

Rezazadeh, A, Jose, S, Durisi, G & Simeone, O 2021, Conditional mutual information-based generalization bound for meta learning. in 2021 IEEE International Symposium on Information Theory (ISIT)., 9518020, IEEE International Symposium on Information Theory proceedings, IEEE, pp. 1176-1181. https://doi.org/10.1109/ISIT45174.2021.9518020

Jose, S & Simeone, O 2021, Information-theoretic bounds on transfer generalization gap based on Jensen-Shannon divergence. in 2021 29th European Signal Processing Conference (EUSIPCO)., 9616270, European Signal Processing Conference , IEEE, pp. 1461-1465, 29th European Signal Processing Conference (EUSIPCO 2021), Dublin, Ireland, 23/08/21. https://doi.org/10.23919/EUSIPCO54536.2021.9616270

Jose, S, Simeone, O & Zhang, Y 2021, Transfer Bayesian meta-learning via weighted free energy minimization. in 2021 IEEE 31st International Workshop on Machine Learning for Signal Processing (MLSP)., 9596239, Proceedings of the IEEE Signal Processing Society Workshop, IEEE, 2021 IEEE 31st International Workshop on Machine Learning for Signal Processing (MLSP), Gold Coast, Queensland, Australia, 20/09/21. https://doi.org/10.1109/MLSP52302.2021.9596239

Jose, S & Simeone, O 2020, Address-event variable-length compression for time-encoded data. in IEEE International Symposium on Information Theory and Applications. <https://ieeexplore.ieee.org/document/9366190>

Jose, S & Kulkarni, AA 2018, Linear Programming Based Finite Blocklength Converses in Information Theory. in IEEE Information Theory and Applications (ITA).

Jose, S & Kulkarni, AA 2018, New finite blocklength converses for asymmetric multiple access channels via linear programming. in IEEE International Conference on Signal Processing and Communications (SPCOM).

View all publications in research portal