Author pages are created from data sourced from our academic publisher partnerships and public sources.
Share This Author
Emulation of physical processes with Emukit
- Andrei Paleyes, Mark Pullin, Maren Mahsereci, Cliff McCollum, Neil D. Lawrence, Javier I. González
- Computer ScienceArXiv
- 25 October 2021
Emukit is a highly adaptable Python toolkit for enriching decision making under uncertainty and allows users to easily prototype new decision making methods for new problems, agnostic to the underlying modeling framework and enables users to use their own custom models.
Challenges in Deploying Machine Learning: a Survey of Case Studies
- Andrei Paleyes, Raoul-Gabriel Urma, Neil D. Lawrence
- Computer ScienceACM Computing Surveys
- 18 November 2020
By mapping found challenges to the steps of the machine learning deployment workflow it is shown that practitioners face issues at each stage of the deployment process.
Causal Bayesian Optimization
It is shown how knowing the causal graph significantly improves the ability to reason about optimal decision making strategies decreasing the optimization cost while avoiding suboptimal solutions.
Good practices for Bayesian Optimization of high dimensional structured spaces
- E. Siivola, Javier I. González, Andrei Paleyes, Aki Vehtari
- Computer ScienceApplied AI Letters
- 31 December 2020
The effect of different search space design choices for performing Bayesian Optimization in high dimensional structured datasets is studied and the influence of the dimensionality of the latent space, the role of the acquisition function and new methods to automatically define the optimization bounds in the latentspace are evaluated.
Automatic Discovery of Privacy–Utility Pareto Fronts
- Brendan Avent, Javier I. González, Tom Diethe, Andrei Paleyes, Borja Balle
- Computer ScienceProc. Priv. Enhancing Technol.
- 26 May 2019
This paper presents a Bayesian optimization methodology for efficiently characterizing the privacy– utility trade-off of any differentially private algorithm using only empirical measurements of its utility.
Effectiveness and resource requirements of test, trace and isolate strategies for COVID in the UK
It is shown that self-isolation of symptomatic individuals and quarantine of their household contacts has a substantial impact on the number of new infections generated by each primary case, and adding contact tracing of non-household contacts of confirmed cases to this broader package of interventions reduces the numberof new infections otherwise generated by 5–15%.
Technical Document 3: E↵ectiveness and Resource Requirements of Test, Trace and Isolate Strategies
This model builds upon the individual-level model of Kucharski et al.
Towards better data discovery and collection with flow-based programming
The potential of flow-based programming (FBP) for simplifying data discovery and collection in software systems and an insight into the current trend that prioritizes model development over data quality management is provided.
An Empirical Evaluation of Flow Based Programming in the Machine Learning Deployment Context
This paper proposes to consider Flow-Based Programming (FBP) as a paradigm for creating DOA applications, and empirically evaluates FBP in the context of ML deployment on four applications that represent typical data science projects, revealing that FBP is a suitable paradigm for data collection and data science tasks.
A penalisation method for batch multi-objective Bayesian optimisation with application in heat exchanger design
HIPPO is shown to be able to cheaply build large batches of informative points by encouraging batch diversity through penalising evaluations with similar predicted objective values, and to demonstrate the application of HIPPO to a challenging heat exchanger design problem, stressing the real-world utility of the highly parallelisable approach to MOBO.