170 research outputs found
Translation and response between Maurice Blanchot and Lydia Davis
Get PDFWhen an author translates a text by another writer, this translation is one form of a response to that text. Other responses may appear in their own writings that are more inflected with their authorial persona. Lydia Davis translated six books by Maurice Blanchot, including fiction and theoretical writings. Blanchot’s concept of the récit privileges non-conventional forms of narrative and it can be considered to have influenced Davis, a view shared in critical writing about Davis. However, responses to his fiction can also be found in Davis’s work. This article reads Lydia Davis’s story “Story” as a response to Maurice Blanchot’s récit, La Folie du jour, translated by Davis as “The Madness of the Day”. Both texts develop a narrative that questions the possibility of arriving at a single story: Blanchot’s narrator cannot tell the story of how he came to have glass ground into his eyes, while Davis’s narrator must try to understand a contradictory story told to her by her lover. However, Davis responds to Blanchot by reversing the perspective in the story: where Blanchot’s narrator must and cannot create a story that explains his situation in a judicial/medical context, Davis’s narrator is struggling to understand her lover’s story which does not explain the situation that they find themselves in. Davis’s narrator is therefore motivated by an emotional need to find an acceptable story that is absent from Blanchot’s narrator. This difference in motivation is central to the difference between Davis’s and Blanchot’s approach, and complicates any reading of his influence on her because she responds to his text in her own
Musculoskeletal Driver Model for the Steering Feedback Controller: Investigating the influence of driving posture on the steering response
No full textHaptic feedback from the steering wheel is one of the most important cues for driver to vehicle interaction. The right feedback is provided by ensuring that the haptic controller provides the required steering feel. Steering feel assessment and design is divided into a subjective and objective approach. The subjective approach entails experiments on the proving ground during which steering parameters can be tuned by steering experts. However, using only subjective assessment is time-consuming, costly and non-repetitive. Since there is no direct method to tune the steering feel objectively, a driver model is required to find a mathematical justification in the mechanical interaction between driver and vehicle during steering. A 3-dimensional multibody arm model is constructed to investigate the influence of driving posture on the nonlinear steering response. It was found that the torque acting in the shoulder joint is higher than in the elbow. The relation between joint torque and joint angles islinear in the shoulder, whereas nonlinearities were found in the elbow joint. Nevertheless, a change of driving posture (i.e. a change of haptic interface) leads to a different steering response. Findings from the driver model were validated by two steering experiments. Muscle contraction was measured in order to analyse the forces acting on the joints.This study shows promise to lead to a different approach for tuning steering parameters. Further investigation and detailed experiments are required to convert this driver model into a method to tune steering feel objectively.Mechanical Engineerin
Proceedings Transborder Library Forum 2007 : bridging the digital divide : crossing all borders = Memorias Foro Transfronterizo de Bibliotecas 2007 : cerrando la brecha digital : cruzando todas las fronteras
No full textIt is with great pleasure that we present this edition of the Proceedings of the Transborder Library Forum (Foro). The 2007 Transborder Library Forum was held at Arizona State University in Tempe, Arizona in February, 2007. We are pleased that there will be both a print edition and an online edition. Editing has been kept to a minimum to preserve the intent of the author in the language the paper was presented. The theme for the 2007 Foro was Bridging the Digital Divide. Topics ranged from international copyright issues to getting information to students in widely dispersed communities with little or no infrastructure except the Internet. While most attendees and speakers were from the USA and Mexico, we also had some from Uganda, Kenya, Hungary, and the West Indies
The impact of reactionary behavior in channel creation games: How actions influence transaction routing in the bitcoin lightning network
No full textPayment channels allow parties to utilize the blockchain to send transactions for a cheaper fee. Previous work has analyzed to which degree a party can profit by facilitating the transaction process. The aim is to increase the usability of the network and to be rewarded for providing this service. However, previous work focuses on maximizing the reward of the individual player in isolation, a model that we aim to expand. That is why in this work we extend the action space to allow other parties to act and react, and observe the impact this has on the rewards of the player that would otherwise act in isolation. Testing existing placement strategies by performing channel placement games, we can assess the difference in the reward that indicates the potential loss that competition may cause when operating in the Bitcoin Lightning Network.Furthermore, we have developed a new strategy that is able to improve the performance in the multi-actor model.Computer Science | Artificial Intelligenc
Permutation-Invariant Tabular Data Synthesis
No full textTabular data synthesis is an emerging approach to circumvent strict regulations on data privacy while discovering knowledge through big data. Although state-of-the-art AI-based tabular data synthesizers, e.g., table-GAN, CTGAN, TVAE, and CTAB-GAN, are effective at generating synthetic tabular data, their training is sensitive to column permutations of input data. In this paper, we first c onduct a n e xtensive e mpirical s tudy to disclose such a property of permutation invariance and an in-depth analysis of the existing synthesizers. We show that changing the input column order worsens the statistical difference between real and synthetic data by up to 38.67% due to the encoding of tabular data and the network architectures. To fully unleash the potential of big synthetic tabular data, we propose two solutions: (i) AE-GAN, a synthesizer that uses an autoencoder network to represent the tabular data and GAN networks to synthesize the latent representation, and (ii) a feature sorting algorithm to find t he s uitable c olumn o rder o f i nput d ata f or CNN-based synthesizers. We evaluate the proposed solutions on five datasets in terms of the sensitivity to the column permutation, the quality of synthetic data, and the utility in downstream analyses. Our results show that we enhance the property of permutation-invariance when training synthesizers and further improve the quality and utility of synthetic data, up to 22%, compared to the existing synthesizers.Green Open Access added to TU Delft Institutional Repository 'You share, we take care!' - Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.Data-Intensive System
FeTGAN: Federated Time-Series Generative Adversarial Network
No full textThe key to producing high-fidelity time-series data is to preserve temporal dynamics. This means that generated sequences respect the relationship between variables across time as in the original data. While new types of GANs have been used to generate time-series data, they, like previous GANimplementations, are time consuming to train. A novel federated framework is proposed, which generates realistic time-series data, by combining supervised and unsupervised training. The framework is based on the work in TimeGAN and Federated GAN (FeGAN). Using an embedded learning space, TimeGANencourages the network to mimic the structure of the training data. FeGAN allows the results of TimeGAN to be combined at a central server, which has benefits for both throughput, and potential to improve data privacy. This also introduces the possibility of using cross domain data. The challenge with creating applying federated learning to TimeGAN, and timeseries data in general is whether the learned temporal dynamics can be combined. This is accomplished by the combination of the weighting and sampling scheme used. This paper demonstrates, by qualitative and quantitative analysis, the ability novel framework proposed, to produce equivalent quality synthetic timeseries data compared to the original TimeGAN, without sharing local data between nodes in the network. This is based on the predictive and discriminative scores described, as well as PCA and t-SNE analysis. Additionally, there is an approximate eleven percent increase in Floating Point Operations per second when using one machine, and up to a thirty percent increase when using multiple.CSE3000 Research ProjectComputer Science and Engineerin
PHD3 loss in cancer enables metabolic reliance on fatty acid oxidation via deactivation of ACC2
No full textWhile much research has examined the use of glucose and glutamine by tumor cells, many cancers instead prefer to metabolize fats. Despite the pervasiveness of this phenotype, knowledge of pathways that drive fatty acid oxidation (FAO) in cancer is limited. Prolyl hydroxylase domain proteins hydroxylate substrate proline residues and have been linked to fuel switching. Here, we reveal that PHD3 rapidly triggers repression of FAO in response to nutrient abundance via hydroxylation of acetyl-coA carboxylase 2 (ACC2). We find that PHD3 expression is strongly decreased in subsets of cancer including acute myeloid leukemia (AML) and is linked to a reliance on fat catabolism regardless of external nutrient cues. Overexpressing PHD3 limits FAO via regulation of ACC2 and consequently impedes leukemia cell proliferation. Thus, loss of PHD3 enables greater utilization of fatty acids but may also serve as a metabolic and therapeutic liability by indicating cancer cell susceptibility to FAO inhibition
From Philosophy of Science to Philosophy of Literature (and Back) via Philosophy of Mind: Philip Kitcher’s Philosophical Pendulum
Get PDFA recent focus of Philip Kitcher’s research has been, somewhat surprisingly in the light of his earlier
work, the philosophical analyses of literary works and operas. Some may see a discontinuity in Kitcher’s
oeuvre in this respect—it may be difficult to see how his earlier contributions to philosophy of science relate
to this much less mainstream approach to philosophy. The aim of this paper is to show that there is no
such discontinuity: Kitcher’s contributions to the philosophy of science and his more recent endeavors into
the philosophy of literature and of music are grounded in the same big picture attitude towards the human
mind—an attitude that he would undoubtedly call ‘pragmatic’: one that emphasizes the importance of those
mental processes that are not (or not entirely) rational.
El análisis filosófico de obras literarias y óperas se ha convertido en un objeto de estudio reciente para
Philip Kitcher, algo quizá sorprendente a la vista de su trabajo anterior. Hay quien puede percibir una discontinuidad
en la obra de Kitcher a este respecto: puede ser difícil apreciar cómo sus anteriores contribuciones
a la filosofía de la ciencia se relacionan con este otro tipo menos mayoritario de filosofía. El propósito
de este artículo es mostrar que no hay tal discontinuidad: las contribuciones de Kitcher a la filosofía de la
ciencia y sus empresas más recientes en filosofía de la literatura y de la música se basan en la misma visión
general del espíritu humano, una actitud que indudablemente él denominaría pragmática: enfatiza la importancia
de los procesos mentales que no son (o no completamente) racionales.
Robust multi-label learning for weakly labeled data
No full textMulti-label learning is one of the hot problems in the field of machine learning. The deep neural networks used to solve it could be quite complex and have a huge capacity. This enormous capacity, however, could also be a negative, as they tend to eventually overfit the undesirable features of the data. One such feature presented in the real-world datasets is imperfect labels. A particularly common type of label imperfection is called weak labels. This corruption is characterized not only by the presence of all relevant labels but also by the addition of some irrelevant ones. In this paper, a novel method, Co-ASL, is introduced to deal with the label noise in multi-label datasets. It combines the state-of-the-art approach for multi-label learning, ASL, with the famous robust training strategy, Co-teaching. The performance of the method is then evaluated on noisy versions of MS-COCO to show the lack of overfitting and the performance improvement over the non-robust multi-label ASL.CSE3000 Research ProjectComputer Science and Engineerin
Robust Multi-label Active Learning for Missing Labels
No full textMulti-label classification has gained a lot of attraction in the field of computer vision over the past couple of years. Here, each instance belongs to multiple class labels simultaneously. There are numerous methods for Multi-label classification, however all of them make the assumption that either the training images are completely labelled or that label correlations are given. Since Active Learning is frequently used when not much data is available, it could be used to determine the missing labels by querying an oracle. This paper proposes a novel solution that combines the current state-of-the-art for Multi-label classification with Active Learning to infer the missing labels. This is done with sampling strategies that try to select the most informative sample from the dataset by exploring the amount of missing labels. With these strategies, we try to minimize the relabeling cost for all samples, while maximizing the information gained. The chosen method called Hard sampling with entropy then looks to select those samples that both the model and we find informative. The chosen measure along with the other measure are then explored and evaluated on a subset of the MSCOCO dataset on 20%, 40% and 60% noise. Hard sampling with entropy then outperforms the state-of-the-art by more then 30%, as well as the baseline sampling method by 2% for 60% noise.CSE3000 Research ProjectComputer Science and Engineerin
- …
