Testing, Learning, and Optimization in High Dimensions by Gatmiry, Khashayar

“…We show that despite nonconvexity, a variant of Stochastic gradient descent (SGD) converges to a good solution for which we prove a novel generalization bound that is proportional to our complexity measure.…”
Get full text

Genera via Deformation Theory and Supersymmetric Mechanics by Wilson, Araminta Amabel

“…Our proof uses the formalism of Gelfand-Kazhdan descent, whose foundations we establish in the super-symplectic setting. …”
Get full text

SGD and Weight Decay Provably Induce a Low-Rank Bias in Deep Neural Networks by Galanti, Tomer, Siegel, Zachary, Gupte, Aparna, Poggio, Tomaso

“…In this paper, we study the bias of Stochastic Gradient Descent (SGD) to learn low-rank weight matrices when training deep ReLU neural networks. …”
Get full text

A framework for analog circuit optimization by Mitros, Piotr, 1979-

Get full text

Model-Based Matching by Linear Combinations of Prototypes by Jones, Michael J., Poggio, Tomaso

“…A stochastic gradient descent algorithm is used to match a model to a novel image by minimizing the error between the model and the novel image. …”
Get full text

DeepMaxSAT : encode logical representation into deep learning models for information extraction by Wu, Meixi

“…To incorporate such logical representation capability to deep learning models, we propose to add a layer of MAX-SAT transformation on top of a deep neural network, which can be trained via end-to-end gradient descent. The integrated model is able to improve task performance under the constraint of logic rules, meanwhile, the weights of the logic rules are adaptable to the training data.…”
Get full text

Energy related activities recognition using smartphones by Tai, Jie Qin

“…By using different types of parameters such as the Adam and Stochastic Gradient Descent (SGD) optimizer, observations are made on how the accuracy of the model will be affected. …”
Get full text

How to be human by Amira Azmi

“…The project invites the audience to reflect on our habit of conformity, ridiculous social conventions and our slow descent into a completely mechanical human life. When provided with information on everything about your life, would you feel relieved by the lack of ambiguity or would you instead have a greater appreciation of your personal freedom?…”
Get full text

Measuring Malayization : kuda kepang and Javanese Muslim Singaporeans in post-independence Singapore (1965-1990) by Sri Amalinah Suhairi

“…Being a Javanese-descent Muslim in post-independence Singapore (1965-1990) is not as straightforward as one thinks. …”
Get full text

Interval forecasting of renewable power generation by Zhang, Hanjie

“…By using the forecast accuracy of extreme learning machine (ELM) and gradient descent, the optimal result of point forecasting is provided, which is the initial data for interval forecasting. …”
Get full text

Cost-sensitive online classification by Hoi, Steven C. H., Wang, Jialei., Zhao, Peilin.

“…In this paper, we formally study this problem, and propose a new framework for cost-sensitive online classification by exploiting the idea of online gradient descent techniques. Based on the framework, we propose a family of cost-sensitive online classification algorithms, which are designed to directly optimize two well-known cost-sensitive measures: (i) maximization of weighted sum of sensitivity and specificity, and (ii) minimization of weighted misclassification cost. …”
Get full text
Get full text

Learning Algorithm effect on Multilayer Feed Forward Artificial Neural Network performance in image coding by Mahmoud, Omer, Anwar, Farhat, Salami, Momoh Jimoh Emiyoka

“…The performance of Multilayer Feed Forward Artificial Neural Network performance in image compression using different learning algorithms is examined in this paper. Based on Gradient Descent, Conjugate Gradient, Quasi-Newton techniques three different error back propagation algorithms have been developed for use in training two types of neural networks, a single hidden layer network and three hidden layers network. …”
Get full text

Learning affordances in object-centric generative models by Wu, Y, Kasewa, S, Groth, O, Salter, S, Sun, L, Parker Jones, O, Posner, H

“…Given only high-level performance criteria (such as task success), accessing these emergent affordances via gradient descent enables the agent to manipulate learned object geometries in a targeted and deliberate way. …”

Convergence to stationarity in the Moran model by Donnelly, P, Rodrigues, E

“…The results depend on a particular duality, and couplings, between X and a genealogical process known as the lines of descent process.…”

Genomic structure and diversity of Plasmodium falciparum in Southeast Asia reveal recent parasite migration patterns by Shetty, A, Jacob, C, Huang, F, Li, Y, Agrawal, S, Saunders, D, Lon, C, Fukuda, M, Ringwald, P, Ashley, E, Han, K, Hlaing, T, Nyunt, M, Silva, J, Stewart, K, Plowe, C, O'Connor, T, Takala-Harrison, S, White, N, Artemisinin Resistance Confirmation, Characterization, and Containment (ARC3), Artemisinin Resistance Containment and Elimination (ARCE), Tracking Resistance to Artemisinin Collaboration (TRAC)

“…Here we elucidate fine-scale parasite population structure and infer recent migration across Southeast Asia using identity-by-descent (IBD) approaches based on genome-wide single nucleotide polymorphisms called in 1722 samples from 54 districts. …”

Concluding remarks: What's in a name? "Negritos" in the context of the human prehistory of South East Asia by Ulijaszek, S

“…The "negrito" hypothesis posits that various indigenous groups throughout Island and Mainland Southeast Asia have a shared phenotype due to common descent from a putative ancestral population, representing a preagricultural substrate of humanity in the region. …”

Racism in the theory canon: Hannah Arendt and 'the one great crime in which America was never involved' by Owens, P

“…However, Arendt also harboured deep racial prejudices, especially when writing about people of African descent, which affected core themes in her political thought. …”

On the Robustness of Visual Homing under Landmark Uncertainty by Schroeter, D, Newman, P

“…In addition, we propose a method that is based on the approximation of the gradient of an implicit objective function, which constitutes a gradient descent approach.We present practical results that show the applicability of our approach using natural visual landmarks, omni-directional vision and a compass. © 2008 The authors and IOS Press. …”

Convergence of policy gradient methods for finite-horizon stochastic linear-quadratic control problems by Giegrich, M, Reisinger, C, Zhang, Y

“…Contrary to discrete-time problems, the cost is noncoercive in the policy and not all descent directions lead to bounded iterates. We propose geometry-aware gradient descents for the mean and covariance of the policy using the Fisher geometry and the Bures-Wasserstein geometry, respectively. …”

Large-scale association analysis provides insights into the genetic architecture and pathophysiology of type 2 diabetes by Morris, A, Voight, B, Teslovich, T, Ferreira, T, Segrè, A, Steinthorsdottir, V, Strawbridge, R, Khan, H, Grallert, H, Mahajan, A, Prokopenko, I, Kang, H, Dina, C, Esko, T, Fraser, R, Kanoni, S, Kumar, A, Lagou, V, Langenberg, C, Luan, J, Lindgren, C, Müller-Nurasyid, M, Pechlivanis, S, Rayner, N, Scott, L

“…To extend understanding of the genetic architecture and molecular basis of type 2 diabetes (T2D), we conducted a meta-analysis of genetic variants on the Metabochip, including 34,840 cases and 114,981 controls, overwhelmingly of European descent. We identified ten previously unreported T2D susceptibility loci, including two showing sex-differentiated association. …”