Global ETD Search
Search theses and dissertations gathered from participating repositories worldwide. Every result links back to the library that holds it. No account is needed.
Results
Showing 1 to 18 of 18 for “"Stochastic optimisation"”.
-
Accelerated Optimisation Algorithms for Machine Learning and Image Processing
… rise in machine learning and image processing. Stochastic optimisation algorithms have long been the methods of choice for training machine learning models. For reasons outlined in this thesis, they are natural choices for solving classical problems in machine learning, and their efficiency …
-
A strategic decision making model on global capacity management for the manufacturing industry under market uncertainty
… under business uncertainty. Conventional optimisation models in this field often suffer from one or more drawbacks, such as deterministic styles, non-inclusive and non-comprehensive decision terms, non-integrated frameworks, non-empirical approaches, small size practices, local/non-global …
-
Differential evolution optimisation of a two echelon inventory system.
… algorithm in the context of spare part inventory optimisation in a two-echelon supply chain setting. The study of DE in application to supply chain planning is limited; thus, we present a well-known hard discrete non-linear stochastic optimisation problem to determine the inventory investment for …
-
Scheduling of flexible manufacturing systems integrating petri nets and artificial intelligence methods.
… processes and Artificial Intelligence (AI) optimisation search techniques. The objective of this integration is to demonstrate its potential in solving a difficult and widely studied problem, the scheduling of Flexible Manufacturing Systems (FIVIS). This work builds on existing results that …
-
On the Co-Optimisation of Reserve Markets: With application to New Zealand
… approach has been extended using a numerical stochastic optimisation model to determine the optimal combined consumption and reserve offer using simulations and dynamic programming. A consumer using this model showed profitable results which approached the theoretical perfect response …
-
Neural Network Training and Inversion with a Bregman Learning Framework
… at the training of DNNs. Built up on distributed optimisation approaches, this work contributes to both the learning problems and the inversion problems of DNNs. In particular, we propose a lifted Bregman learning framework that goes beyond the classical back-propagation approach, and aims to …
-
Scalable Bayesian Inference in the Era of Deep Learning: From Gaussian Processes to Deep Neural Networks
… We address this intractability by using stochastic gradient descent (SGD)---the workhorse algorithm of deep learning---to perform posterior sampling in linear models and their convex duals: Gaussian processes. With this, we turn back to linearised neural networks, finding the linearised …
-
Resolving Graphic Conflict In Scale Reduced Maps: A Simulated Annealing Approach
This thesis explores the use of the stochastic optimisation technique of simulated annealing for cartographic map generalisation. The technique performs operations of displacement, deletion, reduction and enlargement of multiple map objects in order to resolve graphic conflict resulting from a …
-
Smart passive adaptive control of laminated composite plates (through optimisation of fibre orientation)
… Simulated annealing (SA), which is a type of stochastic optimisation method, is used to search for the optimal design. This optimisation algorithm is not based on the starting point and it can escape from the local optimum points. In accordance with the annealing process where temperature …
-
Designing Provably Convergent Algorithms from the Geometry of Data
… we develop algorithms in the context of optimisation and sampling, leveraging the expressive strengths of machine learning while maintaining theoretical interpretations. First using computational imaging as an example application of optimisation, and more generally linear inverse …
-
Design and optimisation of industrial Water Application Planning (WAP): modelling multi-contaminant processes with regeneration technologies
… consumption. In the use of mathematical optimisation methods for solving WAP problems, five major challenges were identified. They are lack of single systematic method to handle reuse/recycle of single and multi-contaminant, a hybrid targeting method which can address multiple local …
-
Investigation of matching problems using constraint programming and optimisation methods
… and techniques to solve them as combinatorial optimisation problems. The first part of the thesis focuses on one-sided matching on a bipartite graph, specifically the popular matching. When the participants express their preferences in a cardinal order, the most common desire is to maximise a …
-
The importance of data collection to support decision making in energy systems
Energy systems need to be designed, operated, and managed effectively to make decarbonisation affordable. But there are many aspects of energy systems, such as their cost and the operating conditions they will face, that are not known precisely at the time these decisions need to be made. To manage …
-
Dynamic voltage optimisation of MV/LV networks
… the potential for implementing dynamic voltage optimisation through CVR on distribution networks in the UK in the presence of high LCT uptake. Studies are conducted for a sample operating year using realistically modelled rural, urban and dense urban distribution networks, representative of …
-
An Optimisation-Based Approach to FKPP-Type Equations
… \tfrac{1}{k} u^{k+1}$ where $\mathcal{L}$ is the stochastic generator of a Markov process, $\phi$ is a function of the space variables and $k\in \mathbb{R}\backslash\{0\}$. An important example, in the case when $k>0$, is equations of the FKPP-type. We also give an example from the theory of …
-
On the complexity of energy landscapes: algorithms and a direct test of the Edwards conjecture
… argue that the approach can be extended to the optimisation of the experimental conditions for observing certain phenomena, for which individual measurements are stochastic and provide little guidance.
-
Verteilungswirkungen anreizorientierter Sozialpolitik : das deutsche Rentenversicherungs- und Steuersystem in der Perspektive dynamischer Lebenszyklusmodelle
Drei große Reformenpakete und eine Reihe kleinerer Begleitmaßnahmen prägen das renten¬politische Erbe der rot-grünen Bundesregierung. Einerseits greifen sie Trends in der Reformpolitik seit Beginn der 90er Jahre auf. So verstärkt die Rentenstrukturreform 2001 beispielsweise die rentenrechtliche …
-
Geometric methods in computational optimal transport and high-dimensional inference
… sets, a dual formulation is derived that enables stochastic updates with O(1) complexity per iteration, independent of support size. Under suitable regularity conditions, non-asymptotic convergence bounds of order O(log(T)/T) for appropriately chosen step-sizes are proved. This framework naturally …