Gaussian Processes for ordinal regression

Problem or Challenge
Highly complex infrastructure projects rarely run to schedule as a result of human optimisation bias during planning, complex interactions between project activities, and failure to learn systematically from past projects. nPlan has amassed the largest proprietary collection of schedules data in the World and developed machine learning modelling technologies that train on historical data to predict future risk. In the real world however, the utility of trained models depends on evaluating well-calibrated probability distributions for predicting future outcomes. The aim of this project is to develop novel probabilistic models that optimise computation of these probability distributions by delineating aleatoric and epistemic contributions to uncertainty using Bayesian inference.
MRes/PhD Project Objectives
1. Develop graph and/or language models for efficient numerical representation of schedule features.
2. Identify representative data sources with similar graph and/or language features to evaluate probabilistic models.
3. Develop probabilistic models that infer relationships between schedule features and outcomes.
4. Optimise modelling algorithms to be scalable for commercial implementation.
PhD Project Description
The development of probabilistic models build upon existing state-of-the-art deep learning technologies already developed at nPlan Since the nPlan project depends on simulating outcomes of schedules based on trained models, it is important that predicted probability distributions provide realistic representations of uncertainty. Since conventional deep learning models do not explicitly characterise probabilities, the aim of the project is to develop novel models that do. Access will be given to the nPlan proprietary data set and the nPlan Core code-base for efficient data reading. The scope of the PhD project is purposefully broad to encourage exploration of different project data types and models, while providing guidance from the expertise within nPlan and the Prof. Girolami group.
MRes Component
The MRes component includes objectives 1&2 listed above. Since the nPlan dataset is very large, preliminary prototype models trained on suitable subsets will be developed within the first year. This work includes identification of data sources with similar structures (Gantt charts represented as directed acyclic graphs) for initial development and evaluation of probabilistic models.

The primary impact of the FIBE2 CDT will be the benefit to society that will accrue from the transformative effect that FIBE2 graduates will have upon current and future infrastructure. The current FIBE CDT has already demonstrated significant impact and FIBE2 will extend this substantially and with particular focus on infrastructure resilience. There will be further impacts across academic research, postgraduate teaching, industry-academia partnering and wider society. Our CDT students are excellent ambassadors and their skills and career trajectories are inspirational. Their outputs so far include >40 journal and conference papers, contributions to a CIRIA report, a book chapter and >15 prizes (e.g. Cambridge Carbon Challenge, EPSRC Doctoral Prizes, best presentation awards). Our students' outreach activities have had far reaching impacts including: Science Festival activities and engineering workshops for school girls. Our innovative CDT training approaches have shifted the culture and priorities in academia and industry towards co-creation for innovation. Our FIBE CDT features in the EPSRC document 'Building Skills for a Prosperous Nation'. Our attention to E&D has resulted in 50% female students with the inspirational ethos attracting students from wide ranging educational backgrounds.

FIBE2 CDT will build on this momentum and expand the scope and reach of our impact. We will capitalise on our major research and training initiatives and strategic collaborations within academia, industry and government to train future infrastructure leaders to address UK and global challenges and this will have direct and significant technical, economic and social impacts for UK infrastructure, its associated stakeholders and civil society at large.

As well as the creation of cohorts of highly skilled research cohorts with cross-disciplinary technical skills, further specific impacts include:

-a transformational cross-disciplinary graduate training and research approach in infrastructure with depth and breadth.

-new forms of Industry-University partnerships. Co-creation with industry of our training and research initiatives has already led to new forms of partnerships such as the I+ scheme, and FIBE2 will further extend this with the 'employer model' variant and others.

-skilled research-minded challenge-focused graduates for UK employers who will derive significant benefit from employing them as catalysts for enterprise, knowledge exchange and innovation, and thus to business growth opportunities.

-enhanced global competitiveness for industrial partners. With our extensive network of 27 industry partners from across all infrastructure sectors who will actively shape the centre with us, we will deliver significant impact and will embrace the cross-disciplinary research emergeing from the CDT to gain competitive advantage.

-support for policy makers at the highest levels of national and local government. The research outcomes and graduates will contribute to an evidence-based foundation for improved decision-making for the efficient management, maintenance and design of infrastructure.

-world-class research outcomes that address national needs, via the direct engagement of our key industrial partners. Other academic institutions will benefit from working with the Centre to collectively advance knowledge.

-wider professional engagement via the creation of powerful informal professional networks between researchers, practitioners, CDT alumni and CDT students, working nationally and internationally, including some hosted by FIBE2 CDT industry partners.

-future generations of infrastructure professional inspired by the FIBE2 CDT's outreach activities whereby pupils, teachers and parents gain insight into the importance of infrastructure engineering.

-the generation of public awareness of the importance of a resilient infrastructure to address inevitable and often unexpected challenges.