Research Proposal
Degree Program: MRes in Mathematical Sciences (Full-Time, 2 Years)
Proposed Title: Mathematical Modelling and Optimization of Dynamic Resource
Allocation Systems
1. Introduction
Efficient resource allocation is vital across numerous industries, including
healthcare, logistics, manufacturing, and IT systems. The mathematical sciences
provide powerful tools for modelling and solving these complex allocation
problems under conditions of uncertainty, variability, and constraint. This
research aims to develop and analyse advanced mathematical models to
optimize resource usage in dynamic environments, contributing both to
theoretical mathematics and real-world applications.
2. Research Aims and Objectives
The overarching aim is to construct mathematical models that can adapt to
dynamic and uncertain conditions in resource allocation systems and to develop
optimization algorithms that improve system performance.
Objectives:
To mathematically formulate resource allocation problems with time-
varying parameters and stochastic inputs.
To investigate and apply deterministic and stochastic optimization
techniques.
To incorporate multi-objective and constraint-based decision models.
To validate models using computational simulations and real-world data,
where available.
To contribute publishable findings to the mathematical sciences
community.
3. Literature Review and Background
Resource allocation problems have historically been studied using tools like
linear and nonlinear programming, queuing theory, and control theory. However,
with the growth in data and system complexity, modern models must also
incorporate uncertainty (using stochastic processes), adaptability (via dynamic
programming), and real-time optimization.
Relevant areas include:
Operations Research: classical optimization techniques and scheduling
algorithms
Stochastic Modelling: Markov decision processes, Poisson processes
Mathematical Programming: multi-objective optimization, constrained
optimization
Simulation Methods: Monte Carlo methods, agent-based modelling
Recent Trends: hybrid models combining machine learning with
optimization
4. Methodology
This research will be approached in four phases across the 2-year period:
Phase 1: Model Development (Months 1–6)
Define and formulate deterministic and stochastic models for resource
allocation.
Explore network-based models, such as flow networks or queuing
networks.
Identify key variables, decision points, constraints, and performance
metrics.
Phase 2: Theoretical Analysis and Algorithm Design (Months 7–12)
Analyse model properties (e.g., convexity, feasibility, duality).
Design solution algorithms using methods such as:
o Integer and linear programming
o Lagrangian relaxation
o Greedy and heuristic methods
o Approximation and decomposition techniques
Phase 3: Simulation and Computational Experiments (Months 13–18)
Implement the models and algorithms using appropriate programming
tools (Python, MATLAB, R).
Create simulated scenarios based on practical cases (e.g., hospital
scheduling, warehouse logistics).
Conduct sensitivity analysis to test robustness and scalability.
Phase 4: Case Study, Publication, and Final Dissertation (Months 19–24)
Apply models to real-world data (via open datasets or partnership if
available).
Evaluate effectiveness and compare with existing methods.
Write and submit at least one academic paper.
Complete final dissertation including all results and conclusions.
5. Significance and Impact
This research is expected to contribute new theoretical insights into dynamic
optimization and practical tools for improving resource efficiency in complex
systems. The resulting models and algorithms may be applicable in domains
such as public service scheduling, supply chain management, and real-time IT
service allocation.
Potential impacts include:
Improved decision-making under uncertainty
More efficient use of limited resources
Academic contributions through publications and conference presentations
6. Timeline (Gantt Summary)
Time
Activity
Frame
In-depth literature review and preliminary model
Months 1–3
setup
Months 4–6 Model formulation and refinement
Months 7–9 Theoretical analysis and algorithm selection
Months 10–
Implementation of optimization algorithms
12
Months 13–
Simulation and numerical testing
15
Months 16–
Validation with real-world or synthetic data
18
Months 19–
Drafting publication and compiling results
21
Months 22–
Dissertation writing and final oral presentation
24
7. Tools and Resources
Software: Python (with NumPy, SciPy, PuLP), MATLAB, R, or Julia
Possible datasets: NHS hospital bed data, UK logistics data, open transport
datasets
Institutional resources: Access to computing clusters or research group
collaboration
8. References
Bertsimas, D., & Tsitsiklis, J. N. (1997). Introduction to Linear Optimization.
Hillier, F. S., & Lieberman, G. J. (2021). Introduction to Operations
Research.
Ross, S. M. (2014). Introduction to Probability Models.
Winston, W. L. (2004). Operations Research: Applications and Algorithms.
