Locally weighted learning methods for non-rigid robot control

This thesis develops a novel approach to robot control that learns to account for a robot's dynamic complexities while executing various control tasks using inspiration from biological sensorimotor control and machine learning. A robot that can learn its own control system can account for complex situations and adapt to changes in control conditions to maximise its performance and reliability in the real world. This research has developed two novel learning methods, with the aim of solving issues with learning control of non-rigid robots that incorporate additional dynamic complexities. The new learning control system was evaluated on a real three degree-of-freedom elastic joint robot arm with a number of experiments: initially validating the learning method and testing its ability to generalise to new tasks, then evaluating the system during a learning control task requiring continuous online model adaptation.

Managing and learning with multiple models: Objectives and optimization algorithms

The quality of environmental decisions should be gauged according to managers' objectives. Management objectives generally seek to maximize quantifiable measures of system benefit, for instance population growth rate. Reaching these goals often requires a certain degree of learning about the system. Learning can occur by using management action in combination with a monitoring system. Furthermore, actions can be chosen strategically to obtain specific kinds of information. Formal decision making tools can choose actions to favor such learning in two ways: implicitly via the optimization algorithm that is used when there is a management objective (for instance, when using adaptive management), or explicitly by quantifying knowledge and using it as the fundamental project objective, an approach new to conservation.This paper outlines three conservation project objectives - a pure management objective, a pure learning objective, and an objective that is a weighted mixture of these two. We use eight optimization algorithms to choose actions that meet project objectives and illustrate them in a simulated conservation project. The algorithms provide a taxonomy of decision making tools in conservation management when there is uncertainty surrounding competing models of system function. The algorithms build upon each other such that their differences are highlighted and practitioners may see where their decision making tools can be improved. © 2010 Elsevier Ltd.

A survey: Error control methods used in bio-cryptography

Integration of biometrics is considered as an attractive solution for the issues associated with password based human authentication as well as for secure storage and release of cryptographic keys which is one of the critical issues associated with modern cryptography. However, the widespread popularity of bio-cryptographic solutions are somewhat restricted by the fuzziness associated with biometric measurements. Therefore, error control mechanisms must be adopted to make sure that fuzziness of biometric inputs can be sufficiently countered. In this paper, we have outlined such existing techniques used in bio-cryptography while explaining how they are deployed in different types of solutions. Finally, we have elaborated on the important facts to be considered when choosing appropriate error correction mechanisms for a particular biometric based solution.

Second-order quantile methods for experts and combinatorial games

We aim to design strategies for sequential decision making that adjust to the difficulty of the learning problem. We study this question both in the setting of prediction with expert advice, and for more general combinatorial decision tasks. We are not satisfied with just guaranteeing minimax regret rates, but we want our algorithms to perform significantly better on easy data. Two popular ways to formalize such adaptivity are second-order regret bounds and quantile bounds. The underlying notions of 'easy data', which may be paraphrased as "the learning problem has small variance" and "multiple decisions are useful", are synergetic. But even though there are sophisticated algorithms that exploit one of the two, no existing algorithm is able to adapt to both. In this paper we outline a new method for obtaining such adaptive algorithms, based on a potential function that aggregates a range of learning rates (which are essential tuning parameters). By choosing the right prior we construct efficient algorithms and show that they reap both benefits by proving the first bounds that are both second-order and incorporate quantiles.

An introduction to education research methods: Exploring the learning journey of pre-service teachers in a transnational programme

Internationally there is interest in developing the research skills of pre-service teachers as a means of ongoing professional renewal with a distinct need for systematic and longitudinal investigation of student learning. The current study takes a unique perspective by exploring the research learning journey of pre-service teachers participating in a transnational degree programme. Using a case-study design that includes both a self-reported and direct measure of research knowledge, the results indicate a progression in learning, as well as evidence that this research knowledge is continued or maintained when the pre-service teachers return to their home university. The findings of this study have implications for both pre-service teacher research training and transnational programmes.

Informing methods for preparing public health overviews of reviews: a comparison of public health overviews with Cochrane Overviews published between 1999 and 2014

Background: Overviews of systematic reviews (SRs) are useful for public health policy; however there is an absence of Cochrane Overviews covering public health (PH) topics. Objectives: We sought to analyze the methodological approaches used in existing Cochrane Overviews and Protocols for overviews (primarily clinical in nature), and compare these to the methods and approaches used in PH overviews (non-Cochrane). The intent was to identify issues that would be relevant for undertaking Cochrane overviews. Methods: We conducted a descriptive analysis of overviews published between 1999 and 2014. We searched the Cochrane Database of Systematic Reviews for Cochrane Protocols for overviews and Cochrane Overviews, and the HealthEvidence.org for PH overviews. The primary characteristics of the overviews and elements of the methodology were extracted and compared. Results: A total of 61 overviews of SRs were included in our analysis; specifically, this included 21 Cochrane Protocols for overviews, 15 Cochrane Overviews, and 27 non-Cochrane PH overviews. Amongst the overviews, the most significant differences are that PH overviews (non-Cochrane) tend to: include earlier and more reviews, greater number of participants, allow lower levels of evidence, use assessment tools other than AMSTAR (A Measurement Tool to Assess Systematic Reviews, i.e. a tool for assessing quality of SRs), not assess quality of evidence in reviews, search more databases overall, specify search limits including English-only reviews, and not consider recent primary studies for inclusion. Some of these differences clearly related to quality, however many relate to the nuances of PH interventions. Conclusions: The methodology in Cochrane overviews and PH overviews varies widely. Future PH overviews may benefit from the Cochrane methodology but the Cochrane approach requires modification to accommodate PH research methodology. Additionally, the use of databases that pre-screen and quality assess relevant PH systematic reviews may help expedite the search process.