No description available.

We address the problem of dynamic user modelling for referring expression generation in spoken dialogue systems, i.e how a spoken dialogue system should choose referring expressions to refer to domain entities to users with different levels of domain expertise, whose domain knowledge is initially unknown to the system. We approach this problem using a statistical planning framework: Reinforcement Learning techniques in Markov Decision Processes (MDP). We present a new reinforcement learning framework to learn user modelling strategies for adaptive referring expression generation (REG) in resource scarce domains (i.e. where no large corpus exists for learning). As a part of the framework, we present novel user simulation models that are sensitive to the referring expressions used by the system and are able to simulate users with different levels of domain knowledge. Such models are shown to simulate real user behaviour more closely than baseline user simulation models. In contrast to previous approaches to user adaptive systems, we do not assume that the user’s domain knowledge is available to the system before the conversation starts. We show that using a small corpus of non-adaptive dialogues it is possible to learn an adaptive user modelling policy in resource scarce domains using our framework. We also show that the learned user modelling strategies performed better in terms of adaptation than hand-coded baselines policies on both simulated and real users. With real users, the learned policy produced around 20% increase in adaptation in comparison to the best performing hand-coded adaptive baseline. We also show that adaptation to user’s domain knowledge results in improving task success (99.47% for learned policy vs 84.7% for hand-coded baseline) and reducing dialogue time of the conversation (11% relative difference). This is because users found it easier to identify domain objects when the system used adaptive referring expressions during the conversations.

Pharmacotherapies for tobacco and alcohol cessation are only modestly successful, so it is important to better understand mechanisms underlying their use and abuse. The overarching goal of this research is to assess a6b2 subunit containing nicotinic acetylcholine receptor (a6b2*nAChR; *denotes possible assembly with other subunits) contributions to abuse-related effects of nicotine and alcohol. In the absence of a6b2*nAChR-selective agonists, a6b2*nAChR gain-of-function (a6L9’S) mice provide a tool for selective activation of a6b2*nAChRs. Using the a6L9’S mice together with nicotine doses sub-threshold for stimulation of native nAChRs, these studies tested the hypothesis that activation of a6b2*nAChRs is sufficient to promote neurochemical and behavioral effects relevant to nicotine addiction. Intracranial infusions of an a6b2*nAChR-selective antagonist further tested the neuroanatomical locus of a6b2*nAChR contributions to mesolimbic dopamine (DA) release and nicotine reward behavior. Our in vivo microdialysis and nicotine conditioned place preference (CPP) studies reveal that stimulation of a6b2*nAChRs on ventral tegmental area (VTA) DA neurons, as well as on DA terminals in the nucleus accumbens (NAc) shell support nicotine reward. VTA a6b2*nAChR stimulation is required for elevated basal NAc DA levels in a6L9’S mice, who also show elevated nicotine CPP. These studies also showed elevated anxiety-like behavior in a6L9’S mice, but no change in a6 subunit null mutant (a6KO) mice to suggest that elevated cholinergic tone at a6b2*nAChRs promotes anxiety-like behavior. To better define the molecular make-up of a6b2*nAChRs supporting nicotine reward and anxiety-like behavior, these studies crossed a6L9’S to a4 subunit knockout mice to differentiate (non-a4)a6b2* and a4a6b2*nAChR contributions. (non-a4)a6b2*nAChRs appear to promote nicotine reward behavior, while the a6b2*nAChR subtype that regulates anxiety-like behavior depends on the anxiety assay. Finally, these studies developed a mouse model of oral operant ethanol (EtOH) self-administration and assessed EtOH reinforcement in a6 heterozygous (a6HET) and a6KO mice to characterize the role of a6b2*nAChRs in EtOH reinforcement. EtOH self-administration was similar to wild type mice in a6KO mice, but not a6HET mice, suggesting that expression of a6b2*nAChRs modulates EtOH reinforcement. Together, these preclinical studies implicate a6b2*nAChRs in various abuse-related effects of nicotine and alcohol, identifying this receptor as a potential therapeutic target for treatment of dependence.

Making good decisions and adapting flexibly to environmental change are critical to the survival of animals. In this thesis, I investigated neural mechanisms underlying suboptimal decision making in humans and underlying behavioural adaptation in monkeys with the use of functional magnetic resonance imaging (fMRI) in both species. In recent decades, in the neuroscience of decision making, there has been a prominent focus on binary decisions. Whether the presence of an additional third option could have an impact on behaviour and neural signals has been largely overlooked. I designed an experiment in which decisions were made between two options in the presence of a third option. A biophysical model simulation made surprising predictions that more suboptimal decisions were made in the presence of a very poor third alternative. Subsequent human behavioural testing showed consistent results with these predictions. In the ventromedial prefrontal cortex (vmPFC), I found that a value comparison signal that is critical for decision making became weaker in the presence of a poor value third option. The effect contrasts with another prominent potential mechanism during multi-alternative decision making – divisive normalization – the signatures of which were observed in the posterior parietal cortex. It has long been thought that the orbitofrontal cortex (OFC) and amygdala mediate reward-guided behavioural adaptation. However, this viewpoint has been recently challenged. I recorded whole brain activity in macaques using fMRI while they performed an object discrimination reversal task over multiple testing sessions. I identified a lateral OFC (lOFC) region in which activity predicted adaptive win-stay/lose-shift behaviour. In contrast, anterior cingulate cortex (ACC) activity predicted future exploratory decisions regardless of reward outcome. Amygdala and lOFC activity was more strongly coupled for adaptive choice shifting and decoupled for task irrelevant reward memory. Day-to-day fluctuations in signals and signal coupling were correlated with day-to-day fluctuations in performance. These data demonstrate OFC, ACC, and amygdala each make unique contributions to flexible behaviour and credit assignment.

This study was undertaken to investigate the facilitating effects of imitation procedures upon a traditional behavior modification program. A "token economy" was instituted within a workshop setting.

This study is concerned with determining the relationship among functional educational level of groups, three types of reinforcement, and length of the study and with their resulting effects on schizophrenic performance.

Avoidance is behaviour that prevents or postpones the occurrence of an aversive stimulus; whereas, escape is behaviour that terminates an aversive stimulus. There are two types of avoidance, both of which prevent an aversive stimulus from occurring. Standard avoidance involves a warning stimulus that predicts an aversive stimulus, reinforced by the termination of the warning stimulus. Alternatively, Sidman or free-operant avoidance does not include a warning stimulus and it is unknown what the exact reinforcing properties are that maintain it. Previous research has found that Betta splendens (Siamese fighting fish) may not engage in Sidman avoidance whereas other fish species including Carassius auratus (goldfish) do. The present study looked at whether Betta splendens could be taught Sidman avoidance when water disturbance is the aversive stimulus, by prompting the fish to emit the avoidance response of crossing over between the two sides of an experimental tank in 30second intervals, and reinforcing the behaviour with a mirror. Results showed an increase in the frequency of independent (i.e., unprompted) crossovers between the sides of the experimental tank resulting in avoidance of water disturbance, an aversive stimulus for these fish. The increase in responding was maintained for one of the three subjects when reinforcement was removed, suggesting that this species can learn Sidman avoidance. / May 2016

The purpose of this series of studies was to evaluate the effects of schedule thinning and delay-to-reinforcement following intervention for individuals diagnosed with intellectual and developmental disabilities. Study one was a systematic review of the available literature on schedule thinning, and study two evaluated the effects of a novel approach to delay-to-reinforcement following functional communication training. Results of both studies found that schedule thinning and delay-to-reinforcement are efficacious procedures for continued reductions in challenging behavior following intervention.

Concrete has many advantages as a low cost and sustainable material. However, more than 5% of the planet’s total carbon emissions are associated with the production of cement, which, in fact, is predominantly due to the large volume of concrete used worldwide. It is known that traditionally designed concrete structures typically use more material than structurally required and, therefore, an important question is whether material demand can be reduced through structural optimisation. A major drawback from optimised design, however, is the cost and complexity of producing conventional rigid moulds. Fabric formwork is emerging as a new method for construction, gaining popularity among architects and engineers for the opportunity to build unique forms and to shape concrete elements efficiently. Porous fabrics, acting as controlled permeability formwork, also have proven effect on the durability characteristics of concrete. While fabric formwork has a profound potential to change the appearance of concrete structures, the shapes cast in fabrics are not defined in advance and have been often created unintentionally. The design of load-bearing reinforced concrete structures, however, requires accurate form-prediction and construction methods for securing steel reinforcement inside flexible fabrics, which presents a number of constructability challenges. For example, cover formers cannot be used to ensure adequate thickness of protective cover, inevitably affecting the acceptance of such structures in practice. This research has demonstrated that non-corrodable FRP reinforcement can be incorporated more easily than steel bars in fabric-formed concrete due to its light weight and flexibility, while it is possible to ensure ductility of such structures through confinement of concrete using FRP helices. A novel splayed anchorage system has been developed to provide end anchorage for optimised sections where standard bends or hooks cannot fit. This work also provides an experimentally verified methodology and guidance for the design and optimisation of fabric-formed elements.

Wong Ngai Ying. / Chinese title in romanization: Zi wo jian cha you qiang hua zuo yong dui jie nan biao xian di ying xiang. / Thesis (M.A.Ed.)--Chinese University of Hong Kong, 1987. / Bibliography: leaves 89-99.