Personal Statement

I have provided Machine-Learning, Software and Data consulting in an ongoing capacity during my Phd. I consulted in the Financial, Recruitment and Property companies across different geographical regions. contact me for data consulting

 

During my Phd, I have actively researched in the areas of scaleable Bayesian methods, large graph cluster counting, graph similarity, information theory, deep learning, random matrix theory, image recognition, natural language processing, optimisation and solution robustness.

 

My main software proffciency is in Python, where I have worked in optimisation and curvature visaulisation for Deep Learning models using Pytorch and previously in Maximum Entropy modelling for graph similarity, graph clustering and Log Determinants. My code is available at Github

 

In my spare time I like to Powerlift and play Music.

Publications

 

Granziol, D., Ru, B., Zohren, S., Dong, X., Osborne, M., and Roberts, S. 2019. MEMe: An accurate maximum entropy method for efficient approximations in large-scale machine learning. Entropy, 21(6), p.551.

 

Granziol, D., and Roberts, S. 2017. Entropic determinants of massive matrices. In 2017 IEEE International Conference on Big Data (Big Data) (pp. 88–93).

 

Granziol et al 2019. Entropic Graph Spectrum. In NeurIPS Workshop on Information Theory and Machine Learning.

 

Granziol et al 2019. How does mini-batching affect curvature for second order deep learning optimization. In NeurIPS Workshop on Beyond First Order Methods in Machine Learning.

 

Fitzsimons, J., Granziol, D., Cutajar, K., Osborne, M., Filippone, M., and Roberts, S. 2017. Entropic trace estimates for log determinants. In Joint European Conference on Machine Learning and Knowledge Discovery in Databases (pp. 323–338).

 

Ru, B., McLeod, M., Granziol, D., and Osborne, M. 2017. Fast information-theoretic Bayesian optimisation. arXiv preprint arXiv:1711.00673.

 

Granziol, D., Wan, X., Garipov, T., Vetrov, D., and Roberts, S. 2019. MLRG Deep Curvature. arXiv preprint arXiv:1912.09656.

 

Granziol, D., Wan, X., and Roberts, S. 2020. Iterate Averaging Helps: An Alternative Perspective in Deep Learning. arXiv preprint arXiv:2003.01247.

 

Diego Granziol, Timur Garipov, Dmitry Vetrov, Stefan Zohren, Stephen Roberts, and Andrew Gordon Wilson. (2020). Towards understanding the true loss surface of deep neural networks using random matrix theory and iterative spectral methods.

Contact Me

If you have any questions

Please get in touch