A comparison of vector symbolic architectures

Vector Symbolic Architectures combine a high-dimensional vector space with a set of carefully designed operators in order to perform symbolic computations with large numerical vectors. Major goals are the exploitation of their representational power and ability to deal with fuzziness and ambiguity. Over the past years, several VSA implementations have been proposed. The available implementations differ in the underlying vector space and the particular implementations of the VSA operators. This paper provides an overview of eleven available VSA implementations and discusses their commonalities and differences in the underlying vector space and operators. We create a taxonomy of available binding operations and show an important ramification for non self-inverse binding operations using an example from analogical reasoning. A main contribution is the experimental comparison of the available implementations in order to evaluate (1) the capacity of bundles, (2) the approximation quality of non-exact unbinding operations, (3) the influence of combining binding and bundling operations on the query answering performance, and (4) the performance on two example applications: visual place- and language-recognition. We expect this comparison and systematization to be relevant for development of VSAs, and to support the selection of an appropriate VSA for a particular task. The implementations are available.

Ansatz-Independent Variational Quantum Classifiers and the Price of Ansatz

The paradigm of variational quantum classifiers (VQCs) encodes classical information as quantum states, followed by quantum processing and then measurements to generate classical predictions. VQCs are promising candidates for efficient utilizations of noisy intermediate scale quantum (NISQ) devices: classifiers involving M-dimensional datasets can be implemented with only ⌈log2 M⌉ qubits by using an amplitude encoding. A general framework for designing and training VQCs, however, is lacking. An encouraging specific embodiment of VQCs, quantum circuit learning (QCL), utilizes an ansatz: a circuit with a predetermined circuit geometry and parametrized gates expressing a time-evolution unitary operator; training involves learning the gate parameters through a gradient- descent algorithm where the gradients themselves can be efficiently estimated by the quantum circuit. The representational power of QCL, however, depends strongly on the choice of the ansatz, as it limits the range of possible unitary operators that a VQC can search over. Equally importantly, the landscape of the optimization problem may have challenging properties such as barren plateaus and the associated gradient-descent algorithm may not find good local minima. Thus, it is critically important to estimate (i) the price of ansatz; that is, the gap between the performance of QCL and the performance of ansatz-independent VQCs, and (ii) the price of using quantum circuits as classical classifiers: that is, the performance gap between VQCs and equivalent classical classifiers. This paper develops a computational framework to address both these open problems. First, it shows that VQCs, including QCL, fit inside the well-known kernel method. Next it introduces a framework for efficiently designing ansatz-independent VQCs, which we call the unitary kernel method (UKM). The UKM framework enables one to estimate the first known bounds on both the price of anstaz and the price of any speedup advantages of VQCs: numerical results with datatsets of various dimensions, ranging from 4 to 256, show that the ansatz-induced gap can vary between 10−20%, while the VQC-induced gap (between VQC and kernel method) can vary between 10−16%. To further understand the role of ansatz in VQCs, we also propose a method of decomposing a given unitary operator into a quantum circuit, which we call the variational circuit realization (VCR): given any parameterized circuit block (as for example, used in QCL), it finds optimal parameters and the number of layers of the circuit block required to approximate any target unitary operator with a given precision.

Performing Darna: The Role of Entertainment Press in Spectacularizing Darna Actresses

This paper looks at the actresses who portrayed Darna and how they are presented as spectacles in the entertainment articles that promote the film and television adaptations. This frame of inquiry comes from the notion that the visual aesthetics of Darna in komiks is largely informed by the superhero genre’s dependence on spectacle as shown in the superhero’s feats of greatness and in her actions and movements which are all larger than life and extraordinary. If this is the case for Darna in print, then how about the actresses tasked with performing her in the movies and television series? How are their bodies being turned into a spectacle in promotional materials in order to conform to the needs of the capital (entertainment industry)? In using the spectacle of the body as framework, the paper also draws on the star system and the role of producers of text in the creation of Darna as we know her today. The aim is to reveal how female bodies were made part of the construction of Darna’s image outside of its fictional universe which results in a discourse that highlights the body of the celebrities rather than Darna’s continuing relevance as a Filipino icon. This sets aside her representational power to embody the struggle and demand of Filipinos for justice and a better life as audience’s attention is diverted towards how these actresses prepared their bodies to perform Darna.

Constructing graphs from genetic encodings

Our understanding of real-world connected systems has benefited from studying their evolution, from random wirings and rewirings to growth-dependent topologies. Long overlooked in this search has been the role of the innate: networks that connect based on identity-dependent compatibility rules. Inspired by the genetic principles that guide brain connectivity, we derive a network encoding process that can utilize wiring rules to reproducibly generate specific topologies. To illustrate the representational power of this approach, we propose stochastic and deterministic processes for generating a wide range of network topologies. Specifically, we detail network heuristics that generate structured graphs, such as feed-forward and hierarchical networks. In addition, we characterize a Random Genetic (RG) family of networks, which, like Erdős–Rényi graphs, display critical phase transitions, however their modular underpinnings lead to markedly different behaviors under targeted attacks. The proposed framework provides a relevant null-model for social and biological systems, where diverse metrics of identity underpin a node’s preferred connectivity.

Towards Improved Medical Image Segmentation Using Deep Learning

Convolutional neural networks have been asserted to be fast and precise frameworks with great potential in image segmentation. Within the medical domain, image segmentation is a pre-cursor to several applications including surgical simulations, treatment planning and patient prognosis. In this thesis, we attempt to solve two major limitations of current segmentation practices: 1) dealing with unbalanced classes and 2) dealing with multiple modalities. In medical imaging, unbalanced classes present as the regions of interest that are typically significantly smaller in volume than the background class or other classes. We propose an improvement to the current gold standard cost function to boost the focus of the network to the smaller classes. Another problem within medical imaging is the variation in both anatomy and pathology across patients. Utilizing multiple imaging modalities provides complementary, segmentation-specific information and is commonly employed by radiologists when contouring data. We propose a image fusion strategy for multi-modal data that uses the variation in modality specific features to guide the task specific learning. Together, our contributions propose a framework to maximize the representational power of the dataset using models with less complexity and higher generalizability. Our contributions outperform baseline models for multi-class segmentation and are modular enough to be scaled up to deeper networks. We demonstrate the effectiveness of the proposed cost function and multimodal framework, both individually and together, on benchmark datasets including the Breast Ultrasound Dataset B (BUS) [1], the International Skin Imaging Collaboration (ISIC 2018) [2], [3] and the Brain Tumor Segmentation Challenge (BraTs 2018) [4]. In all experiments, the proposed methods match or outperform the baseline methods while employing simpler networks

Towards Improved Medical Image Segmentation Using Deep Learning

Convolutional neural networks have been asserted to be fast and precise frameworks with great potential in image segmentation. Within the medical domain, image segmentation is a pre-cursor to several applications including surgical simulations, treatment planning and patient prognosis. In this thesis, we attempt to solve two major limitations of current segmentation practices: 1) dealing with unbalanced classes and 2) dealing with multiple modalities. In medical imaging, unbalanced classes present as the regions of interest that are typically significantly smaller in volume than the background class or other classes. We propose an improvement to the current gold standard cost function to boost the focus of the network to the smaller classes. Another problem within medical imaging is the variation in both anatomy and pathology across patients. Utilizing multiple imaging modalities provides complementary, segmentation-specific information and is commonly employed by radiologists when contouring data. We propose a image fusion strategy for multi-modal data that uses the variation in modality specific features to guide the task specific learning. Together, our contributions propose a framework to maximize the representational power of the dataset using models with less complexity and higher generalizability. Our contributions outperform baseline models for multi-class segmentation and are modular enough to be scaled up to deeper networks. We demonstrate the effectiveness of the proposed cost function and multimodal framework, both individually and together, on benchmark datasets including the Breast Ultrasound Dataset B (BUS) [1], the International Skin Imaging Collaboration (ISIC 2018) [2], [3] and the Brain Tumor Segmentation Challenge (BraTs 2018) [4]. In all experiments, the proposed methods match or outperform the baseline methods while employing simpler networks

Frozen Out: Audiences, Affect and Women’s Hockey

Every four years, millions of Canadians watch women play hockey during the Olympics. Yet when it comes to regularly scheduled professional games, that audience dramatically decreases. In 2019, low audience numbers led to the closure of the Canadian Women’s Hockey League and put the future of professional women’s hockey in jeopardy. As with many women’s sports, broadcasters argue the cost of production is too great, the value of airtime minutes too high to take the financial risk of televising the women’s game without guaranteeing viewers for advertisers. Activists and athletes argue that the audience must be built through broadcaster investment. While scholars have examined hockey for its representational power to define national and gendered identities, there has been shockingly little research into the hockey audience. This mixed method audience reception study seeks to explore the viewing inconsistencies of the audience for women’s hockey. Quantitative results from an online survey (n =685) provided data about viewing habits, perceptions and knowledge. This data informed qualitative focus groups (n = 25) that in turn provided contextualization and reasoning for the quantitative data. Mixed method analysis intersected grounded theory with audience reception, sport media, feminist studies and affect theory to identify a persistent discursive strategy framing women’s hockey as “pure” for resisting the crass commercialization, incessant violence and individualistic star system of professional men’s hockey. I argue that women’s hockey becomes the manifestation of the Canadian myth of hockey; men’s hockey as it used to be. As a nostalgic placeholder devoid of context, contemporary women’s hockey functions within a double bind; virtuous and elevated yet non-viable as a commercial enterprise. This ensures that the sport remains precarious at best. “Pure” women’s hockey also functions as a postfeminist essentializing discourse that solves the gender risk of hockey’s hypermasculinity while disavowing women’s physically aggressive play and sport media’s affective currency. Whereas Olympic women’s hockey relies on patriotic pride for audience affective engagement, professional women’s hockey is framed by cognitive contradictions, “pure” but commercial, gender normative but transgressive. Confused and disconnected from the game and players, audiences are left unaffected.

Frozen Out: Audiences, Affect and Women’s Hockey

Every four years, millions of Canadians watch women play hockey during the Olympics. Yet when it comes to regularly scheduled professional games, that audience dramatically decreases. In 2019, low audience numbers led to the closure of the Canadian Women’s Hockey League and put the future of professional women’s hockey in jeopardy. As with many women’s sports, broadcasters argue the cost of production is too great, the value of airtime minutes too high to take the financial risk of televising the women’s game without guaranteeing viewers for advertisers. Activists and athletes argue that the audience must be built through broadcaster investment. While scholars have examined hockey for its representational power to define national and gendered identities, there has been shockingly little research into the hockey audience. This mixed method audience reception study seeks to explore the viewing inconsistencies of the audience for women’s hockey. Quantitative results from an online survey (n =685) provided data about viewing habits, perceptions and knowledge. This data informed qualitative focus groups (n = 25) that in turn provided contextualization and reasoning for the quantitative data. Mixed method analysis intersected grounded theory with audience reception, sport media, feminist studies and affect theory to identify a persistent discursive strategy framing women’s hockey as “pure” for resisting the crass commercialization, incessant violence and individualistic star system of professional men’s hockey. I argue that women’s hockey becomes the manifestation of the Canadian myth of hockey; men’s hockey as it used to be. As a nostalgic placeholder devoid of context, contemporary women’s hockey functions within a double bind; virtuous and elevated yet non-viable as a commercial enterprise. This ensures that the sport remains precarious at best. “Pure” women’s hockey also functions as a postfeminist essentializing discourse that solves the gender risk of hockey’s hypermasculinity while disavowing women’s physically aggressive play and sport media’s affective currency. Whereas Olympic women’s hockey relies on patriotic pride for audience affective engagement, professional women’s hockey is framed by cognitive contradictions, “pure” but commercial, gender normative but transgressive. Confused and disconnected from the game and players, audiences are left unaffected.