The Score Function Estimator Is Widely Used For Estimating Gradients Of Stochastic Objectives In Stochastic Computation Graphs (scg), Eg. In Reinforcement Learning And Meta-learning. While Deriving The First-order Gradient Estimators By Differentiating A Surrogate Loss (sl) Objective Is Computationally And Conceptually Simple, Using The Same Approach For Higher-order Gradients Is More Challenging. Firstly, Analytically Deriving And Implementing Such Estimators Is Laborious And Not Compliant With Automatic Differentiation. Secondly, Repeatedly Applying Sl To Construct New Objectives For Each Order Gradient Involves Increasingly Cumbersome Graph Manipulations. Lastly, To Match The First-order Gradient Under Differentiation, Sl Treats Part Of The Cost As A Fixed Sample, Which We Show Leads To Missing And Wrong Terms For Higher-order Gradient Estimators. To Address All These Shortcomings In A Unified Way, We Introduce Dice, Which Provides A Single Objective That Can Be Differentiated Repeatedly, Generating Correct Gradient Estimators Of Any Order In Scgs. Unlike Sl, Dice Relies On Automatic Differentiation For Performing The Requisite Graph Manipulations. We Verify The Correctness Of Dice Both Through A Proof And Through Numerical Evaluation Of The Dice Gradient Estimates. We Also Use Dice To Propose And Evaluate A Novel Approach For Multi-agent Learning. Our Code Is Available At Https://goo.gl/xkkgxn.
A Key Challenge In Complex Visuomotor Control Is Learning Abstract Representations That Are Effective For Specifying Goals, Planning, And Generalization. To This End, We Introduce Universal Planning Networks (upn). Upns Embed Differentiable Planning Within A Goal-directed Policy. This Planning Computation Unrolls A Forward Model In A Latent Space And Infers An Optimal Action Plan Through Gradient Descent Trajectory Optimization. The Plan-by-gradient-descent Process And Its Underlying Representations Are Learned End-to-end To Directly Optimize A Supervised Imitation Learning Objective. We Find That The Representations Learned Are Not Only Effective For Goal-directed Visual Imitation Via Gradient-based Trajectory Optimization, But Can Also Provide A Metric For Specifying Goals Using Images. The Learned Representations Can Be Leveraged To Specify Distance-based Rewards To Reach New Target States For Model-free Reinforcement Learning, Resulting In Substantially More Effective Learning When Solving New Tasks Described Via Image-based Goals. We Were Able To Achieve Successful Transfer Of Visuomotor Planning Strategies Across Robots With Significantly Different Morphologies And Actuation Capabilities.
The purpose of this paper is to analyse data on first-year students’ needs regarding academic support services and reasons for their intention to leave the institution prior to degree completion.
Vol 1 No 2 (2014): Selected and Extended Papers from the Third International Conference on Learning Analytics & Knowledge Special Issue - LAK'13 Selected, Extended, and Revised Papers
W. Hamilton, R. Ying, and J. Leskovec. (2017)cite arxiv:1709.05584Comment: Published in the IEEE Data Engineering Bulletin, September 2017; version with minor corrections.
A. Chéritat. (2014)cite arxiv:1410.4417Comment: 16 pages, 7 figures. This version has the following changes: Added computer generated images of the key positions S1 and S2. Corrected several minor mistakes. Corrected the proof of the main proposition (I had forgotten to ensure that the top and bottom curves remain embedded during the homotopy) and slightly changed the statement of Lemma 3 to adapt.
P. Huang, K. Matzen, J. Kopf, N. Ahuja, and J. Huang. (2018)cite arxiv:1804.00650Comment: CVPR 2018. Project page: https://phuang17.github.io/DeepMVS/ Code: https://github.com/phuang17/DeepMVS.
S. Albrecht, and P. Stone. (2017)cite arxiv:1709.08071Comment: 42 pages, submitted for review to Artificial Intelligence Journal. Keywords: multiagent systems, agent modelling, opponent modelling, survey, open problems.
R. Sharipov. (2002)cite arxiv:cs/0201007Comment: AmSTeX, 7 pages, amsppt style, English wording is improved, references are transformed to hyperlinks, the fugure is incorporated into the PS and PDF files.
W. Hung, Y. Tsai, Y. Liou, Y. Lin, and M. Yang. (2018)cite arxiv:1802.07934Comment: Accepted in BMVC 2018. Code and models available at https://github.com/hfslyc/AdvSemiSeg.
G. Reinmann. (Arbeitsbericht Nr. 18). Augsburg: Universität Augsburg, Medienpädagogik. URL: http://www.imb-uni-augsburg.de/files/Arbeitsbericht_18.pdf, (2008)
J. Schmidt, S. Dreyer, and C. Lampert. Arbeitspapiere des Hans Bredow-Instituts Nr. 19 Hamburg: Verlag Hans-Bredow-Institut 2008, URL: http://www.hans-bredow-institut.de/publikationen/apapiere/19Onlinespiele.pdf, (June 2008)
G. Reinmann-Rothmeier. Forschungsbericht Nr. 129. München: Ludwig-Maximilians-Universität, Lehrstuhl für Empirische Pädagogik und Pädagogische Psychologie. URL: http://epub.ub.uni-muenchen.de/237/1/FB_129.pdf, (Dezember 2000)
A. Alemi, and I. Fischer. (2018)cite arxiv:1807.04162Comment: Presented at the ICML 2018 workshop on Theoretical Foundations and Applications of Deep Generative Models.
G. Reinmann, and T. Bianco. (Arbeitsbericht Nr. 17). Augsburg: Universität Augsburg, Medienpädagogik. URL: http://www.imb-uni-augsburg.de/files/Arbeitsbericht_17.pdf, (February 2008)
M. Braun, S. Krebs, F. Flohr, and D. Gavrila. (2018)cite arxiv:1805.07193Comment: Submitted to IEEE Trans. on Pattern Analysis and Machine Intelligence.
A. Jameson, O. Jacobs, and B. Steimel. Ein Projekt von mind Business Consultants und Strateco in Zusammenarbeit mit DFKI (Herausgeber NürnbergMesse GmbH), (September 2009)Studie anforderbar unter http://www.mind-consult.net/.
D. Sonntag, R. Neßelrath, G. Sonnenberg, and G. Herzog. Paper presented at the First International Workshop on Spoken Dialogue Systems Technology (IWSDS-2009), Kloster Irsee, Germany, (December 2009)Available from http://www.dfki.de/web/forschung/publikationen?pubid=4673.
A. Mousavian, D. Anguelov, J. Flynn, and J. Kosecka. (2016)cite arxiv:1612.00496Comment: To appear in IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2017.