Classical quantifiers (e.g., “all”, “some” and “none”) have been extensively studied in logic and psychology. In contrast, generalized quantifiers (e.g., “most”) allow for fine-grained statements about quantities. The discrepancy in the underlying mental representation and its interpretation among interpreters can affect language use and reasoning. We investigated the effect of quantifier type, quantification space (set size) and monotonicity on processing difficulty (in response time, RT) and response diversity of 77 generalized quantifiers. Shannon entropy was employed to measure response diversity. Our findings indicate: (i) Set size is a significant factor of response diversity, which implies that the underlying space is relevant for the interpretation. (ii) Quantifiers possess a rather static underlying representation within and across tasks within a participant. (iii) Quantifier type and monotonicity can affect response diversity; while the response diversity can predict RT. (iv) In reasoning, the number of generalized quantifiers versus classical quantifiers in a syllogism is a factor of re- sponse diversity. Diversity in the interpretation of generalized quantifiers may be a cause of human’s deviation from logical responses.

Plasmonics, a sub-field of nanophotonics, has attracted amazing interests in recent years due to its promise of revolutionary impacts on nanotechnologies, medicine, and our daily lives. In short, plasmonics guides, confines, localizes, and enhances electromagnetic waves to a small area to enable novel applications such as plasmonic waveguides, sub-wavelength diffraction lithography and imaging, higher photovoltaic efficiency, sub-wavelength lasers, and biological sensing. The first part of this dissertation will introduce the general concepts of plasmonics and its applications in biological sensing, starting from Maxell's equations and the Drude model of metal. In the second part, it will delve into the specifics of designing, simulating, fabricating, and characterizing nanostructures that show improved performance characteristics over previous published results. Plasmonic coupling of propagating surface plasmon polaritons and localized plasmons is utilized to fabricate three-dimensional mushroom-like metallodielectric array for improved surface plasmon resonance biological sensing over conventional nanohole array. It is used to determine the hydrophilicity of a surface by monitoring the shift in resonant wavelength. In addition, the fabricated structure is used to detect specific binding events between biotinylated-bovine serum albumin and streptavidin protein and demonstrates an improved limit of detection. The real- time control and manipulation of nanoparticles in a biocompatible aqueous solution is demonstrated in a multilayer micro/nanofluidic chip platform by using voltage control. This will have future applications to control and position nanoparticles and biomolecules to an area of maximum field enhancement to co-localize the analyte with the plasmonic nanostructure for maximal enhancement efficiency. Finally, the intra-particle plasmonic coupling of a nanotorch structure is utilized for surface enhanced Raman scattering detection to yield high enhancement factor and better than 80\% Raman signal uniformity due to the fabrication of controllable rim opening of the nanotorch and the deterministic orientation of the structure. The common theme of this dissertation is coupled plasmonic nanostructures for improved biological sensing applications

A novel noncontact indentation system with the combination of an air jet and optical coherence tomography (OCT) was presented in this paper for the quantitative measurement of the mechanical properties of soft tissues. The key idea of this method is to use a pressure-controlled air jet as an indenter to compress the soft tissue in a noncontact way and utilize the OCT signals to extract the deformation induced. This indentation system provides measurement and mapping of tissue elasticity for small specimens with high scanning speed. Experiments were performed on 27 silicone tissue-mimicking phantoms with different Young's moduli, which were also measured by uniaxial compression tests. The regression coefficient of the indentation force to the indentation depth (N mm(-1)) was used as an indicator of the stiffness of tissue under air jet indentation. Results showed that the stiffness coefficients measured by the current system correlated well with the corresponding Young's moduli obtained by conventional mechanical testing (r = 0.89, p < 0.001). Preliminary in vivo tests also showed that the change of soft tissue stiffness with and without the contraction of the underlying muscles in the hand could be differentiated by the current measurement. This system may have broad applications in tissue assessment and characterization where alterations of mechanical properties are involved, in particular with the potential of noncontact micro-indentation for tissues.

This music score was submitted for the Kaleidoscope 2020 Call for Scores, an open access collaboration with the UCLA Music Library.

A wireless mesh network (WMN) is a hierarchical network consisting of mesh clients, mesh routers and gateways. Mesh routers constitute a wireless mesh backbone, to which mesh clients are connected, and gateways are chosen among mesh routers providing Internet access. In this dissertation, throughput performance of WMNs has been investigated from theoretical capacity analysis to practical algorithms' design. For such a network with Nc randomly distributed mesh clients, regularly placed mesh routers and Ng gateways, assuming that each mesh router can transmit at W bits/s, the per-client throughput capacity has been derived as a function of Nc, Nr, Ng and W. The result illustrates that, in order to achieve high capacity performance, mesh backbone network must be optimally designed. An innovative gateway placement algorithm is thus developed. It determines the location of a gateway based on a new performance metric. Therefore, given a certain number of gateways, the proposed gateway placement scheme provides a framework of maximizing the throughput of WMNs through proper placement of these gateways. Experimental results show that it constantly outperforms other schemes with a large margin. Algorithms on choosing the optimal number of mesh routers are also presented. Two problems are investigated. In Maximum Throughput Partition problem, the ideal throughput is achieved by optimally partitioning the network with a proper number of backbone nodes. In Maximum Throughput Partition with Hops' number Constraint problem, a similar problem is studied but with constraint on the average number of hops in the backbone network. The results show that it is critical to find an appropriate size of the backbone network for a WMN, especially when the hops' number constraint is imposed. The research findings and results can be used as guidelines for protocol design and deployment of WMNs

This study explores the self-organizing neural network as a model of lexical and morphological acquisition. We examined issues of generalization, representation, and recovery in a multiple feature-map model. Our results indicate that self-organization and Hebbian learning are two important computational principles that can account for the psycholinguistic processes of semantic representation, morphological generalization, and recovery from generalizations in the acquisition of reversive prefixes such as un- and dis-. These results attest to the utility of self-organizing neural networks in the study of language acquisition.