Dynamic-Domain RRTs. Anna Yershova 1 Léonard Jaillet 2 Thierry Siméon 2 Steven M. LaValle 1

Size: px
Start display at page:

Download "Dynamic-Domain RRTs. Anna Yershova 1 Léonard Jaillet 2 Thierry Siméon 2 Steven M. LaValle 1"

Transcription

1 Dynamic-Domain RRTs Anna Yershova 1 Léonard Jaillet 2 Thierry Siméon 2 Steven M. LaValle 1 1. University of Illinois at Urbana-Champaign 2. LAAS/CNRS Toulouse To appear: ICRA 2005 Barcelona Thanks to: US National Science Foundation, UIUC/CNRS funding, Kineo

2 Rapidly-exploring Random Trees (RRTs Introduced by LaValle and Kuffner, ICRA Applied, adapted, and extended in many works: Frazzoli, Dahleh, Feron, 2000; Toussaint, Basar, Bullo, 2000; Vallejo, Jones, Amato, 2000; Strady, Laumond, 2000; Mayeux, Simeon, 2000; Karatas, Bullo, 2001; Li, Chang, 2001; Kuffner, Nishiwaki, Kagami, Inaba, Inoue, 2000, 2001; Williams, Kim, Hofbaur, How, Kennell, Loy, Ragno, Stedl, Walcott, 2001; Carpin, Pagello, 2002; Branicky, Curtiss, 2002; Cortes, Simeon, 2004; Urmson, Simmons, 2003; Yamane, Kuffner, Hodgins, 2004; Strandberg, 2004;... Also, applications to biology, computational geography, verification, virtual prototyping, architecture, solar sailing, computer graphics,...

3 The RRT Construction Algorithm BUILD RRT(q init ) 1 T.init(q init ); 2 for k = 1 to K do 3 q rand RANDOM CONFIG(); 4 EXTEND(T, q rand ); EXTEND(T, q rand ) ε q new q init q near q Metric on C: ρ : C C [0, ) Nearest neighbors: Atramentov [Yershova], LaValle, 2002; Arya, Mount, 1997;... Incremental collision detection: Lin, Canny, 1991; Mirtich, 1997

4 A Rapidly-Exploring Random Tree (RRT)

5 Voronoi-Biased Exploration

6 Voronoi Diagram in R2

7 Voronoi Diagram in R2

8 Voronoi Diagram in R2

9 Refinement vs. Expansion Refinement Expansion Where will the random sample fall?

10 Limit Case: Pure Expansion Let X be an n-dimensonal ball, X = {x R n x r}, in which r is very large. The RRT will explore n + 1 opposing directions. The principle directions are vertices of a regular (n + 1)-simplex:

11 Determining the Boundary Expansion dominates Balanced refiniment and expansion The tradeoff depends on the size of the bounding box.

12 When is Expanson Bad? Refinement is good because because it is like multiresolution search. Expansion is bad when blocked by obstacles.

13 Bug Trap Small Bouding Box Large Bounding Box Which one will perform better?

14 Voronoi Bias for the Original RRT

15 Visibility-Based Clipping of the Voronoi Regions Nice idea, but how can this be done in practice? Even better: Voronoi diagram for obstacle-based metric

16 A Boundary Node Cobst (a) Cobst (b) Cobst (c) v v v (a) Regular RRT, unbounded Voronoi region (b) Visibility region (c) Dynamic domain

17 A Non-Boundary Node (a) (b) (c) C obst C obst C obst v v v R (a) Regular RRT, unbounded Voronoi region (b) Visibility region (c) Dynamic domain

18 Dynamic-Domain RRT Bias

19 DD-RRT Algorithm BUILD DYNAMIC DOMAIN RRT(q init ) 1 T.init(q init ); 2 for k = 1 to K do 3 repeat 4 q rand RANDOM CONFIG(); 5 q near NEAREST NEIGHBOR(q rand, T ); 6 until dist(q near, q rand ) < q near.radius 7 if CONNECT(T, q rand, q near, q new ) 8 q new.radius = ; 9 T.add vertex(q new ); 10 T.add edge(q near, q new ); 11 else 12 q near.radius = R; 13 Return T ;

20 Implementation Details MOVE3D (LAAS/CNRS) 333 Mhz Sunblade 100 with SunOs 5.9 (not very fast) Compiler: GCC 3.3 Fast nearest neighbor searching (Yershova [Atramentov], LaValle, 2002)

21 Experiments Two kinds: Controlled experiments for toy problems Challenging benchmarks from industry and biology

22 Shrinking Bug Trap (1) (2) (3) Large Medium Small

23 Shrinking Bug Trap Trap Size Statistic Dynamic-Domain bi-rrt bi-rrt Large time (1) 0.4 sec 0.1 sec no. nodes (1) CD calls (1) Medium time (2) 2.5 sec 379 sec no. nodes (2) CD calls (2) Small time (3) 1.6 sec > sec no. nodes (3) 1301 CD calls (3) 3022 The smaller the bug trap, the better the improvement.

24 (Example provided courtesy of KINEO) Wiper Motor

25 Wiper Motor Dynamic-Domain bi-rrt bi-rrt time 217 sec > sec no. nodes 219 CD calls [Movie]

26 Molecule

27 Molecule Dynamic-Domain bi-rrt bi-rrt time 70 sec 2926 sec no. nodes CD calls Improvement factor: 41.8 DD-RRT also solved in one hour a conformational search problem with 330 DOFs!

28 Labyrinth

29 Labyrinth Dynamic-Domain bi-rrt bi-rrt time 161 sec 237 sec no. nodes CD calls [Movie]

30 Conclusions Balancing the amount of refinement and expansion is important in RRTs. Provides dramatic performing improvements on some problems. Does not incur much penalty for unsuitable problems. Work in Progress: There is a radius parameter. Adaptive tuning is possible. Still investgating several variatons of the algorithm. Application to planning under differential constrants. Application to planning for closed chains.

Robotics. Path Planning. Marc Toussaint U Stuttgart

Robotics. Path Planning. Marc Toussaint U Stuttgart Robotics Path Planning Path finding vs. trajectory optimization, local vs. global, Dijkstra, Probabilistic Roadmaps, Rapidly Exploring Random Trees, non-holonomic systems, car system equation, path-finding

More information

Robotics. Path Planning. University of Stuttgart Winter 2018/19

Robotics. Path Planning. University of Stuttgart Winter 2018/19 Robotics Path Planning Path finding vs. trajectory optimization, local vs. global, Dijkstra, Probabilistic Roadmaps, Rapidly Exploring Random Trees, non-holonomic systems, car system equation, path-finding

More information

Control of probabilistic diffusion in motion planning

Control of probabilistic diffusion in motion planning Control of probabilistic diffusion in motion planning Sébastien Dalibard and Jean-Paul Laumond LAAS-CNRS University of Toulouse {sdalibar,jpl}@laas.fr Abstract: The paper presents a method to control probabilistic

More information

Completeness of Randomized Kinodynamic Planners with State-based Steering

Completeness of Randomized Kinodynamic Planners with State-based Steering Completeness of Randomized Kinodynamic Planners with State-based Steering Stéphane Caron 1, Quang-Cuong Pham 2, Yoshihiko Nakamura 1 Abstract The panorama of probabilistic completeness results for kinodynamic

More information

TWo decades ago LaValle and Kuffner presented the

TWo decades ago LaValle and Kuffner presented the IEEE ROBOTICS AND AUTOMATION LETTERS. PREPRINT VERSION. ACCEPTED NOVEMBER, 2018 1 Probabilistic completeness of RRT for geometric and kinodynamic planning with forward propagation Michal Kleinbort 1, Kiril

More information

Programming Robots in ROS Slides adapted from CMU, Harvard and TU Stuttgart

Programming Robots in ROS Slides adapted from CMU, Harvard and TU Stuttgart Programming Robots in ROS Slides adapted from CMU, Harvard and TU Stuttgart Path Planning Problem Given an initial configuration q_start and a goal configuration q_goal,, we must generate the best continuous

More information

Enhancing sampling-based kinodynamic motion planning for quadrotors

Enhancing sampling-based kinodynamic motion planning for quadrotors Enhancing sampling-based kinodynamic motion planning for quadrotors Alexandre Boeuf, Juan Cortés, Rachid Alami, Thierry Siméon To cite this version: Alexandre Boeuf, Juan Cortés, Rachid Alami, Thierry

More information

Test Coverage for Continuous and Hybrid Systems

Test Coverage for Continuous and Hybrid Systems Test Coverage for Continuous and Hybrid Systems Tarik Nahhal and Thao Dang VERIMAG,2avenuedeVignate 38610 Gières, France Abstract. We propose a novel test coverage measure for continuous and hybrid systems,

More information

Motion Planning with Invariant Set Trees

Motion Planning with Invariant Set Trees MITSUBISHI ELECTRIC RESEARCH LABORATORIES http://www.merl.com Motion Planning with Invariant Set Trees Weiss, A.; Danielson, C.; Berntorp, K.; Kolmanovsky, I.V.; Di Cairano, S. TR217-128 August 217 Abstract

More information

Probabilistically Complete Planning with End-Effector Pose Constraints

Probabilistically Complete Planning with End-Effector Pose Constraints Probabilistically Complete Planning with End-Effector Pose Constraints Dmitry Berenson Siddhartha S. Srinivasa The Robotics Institute Intel Labs Pittsburgh Carnegie Mellon University 4720 Forbes Ave.,

More information

Completeness of Randomized Kinodynamic Planners with State-based Steering

Completeness of Randomized Kinodynamic Planners with State-based Steering Completeness of Randomized Kinodynamic Planners with State-based Steering Stéphane Caron 1, Quang-Cuong Pham, Yoshihiko Nakamura 1 Abstract The panorama of probabilistic completeness results for kinodynamic

More information

Supplementary Material for the paper Kinodynamic Planning in the Configuration Space via Velocity Interval Propagation

Supplementary Material for the paper Kinodynamic Planning in the Configuration Space via Velocity Interval Propagation Supplementary Material for the paper Kinodynamic Planning in the Configuration Space via Velocity Interval Propagation Quang-Cuong Pham, Stéphane Caron, Yoshihiko Nakamura Department of Mechano-Informatics,

More information

Improving the Performance of Sampling-Based Planners by Using a Symmetry-Exploiting Gap Reduction Algorithm

Improving the Performance of Sampling-Based Planners by Using a Symmetry-Exploiting Gap Reduction Algorithm Improving the Performance of Sampling-Based Planners by Using a Symmetry-Exploiting Gap Reduction Algorithm Peng Cheng Emilio Frazzoli Steven M. LaValle University of Illinois Urbana, IL 61801 USA {pcheng1,

More information

Localization aware sampling and connection strategies for incremental motion planning under uncertainty

Localization aware sampling and connection strategies for incremental motion planning under uncertainty Autonomous Robots manuscript No. (will be inserted by the editor) Localization aware sampling and connection strategies for incremental motion planning under uncertainty Vinay Pilania Kamal Gupta Received:

More information

arxiv: v2 [cs.ro] 20 Nov 2015

arxiv: v2 [cs.ro] 20 Nov 2015 Completeness of Randomized Kinodynamic Planners with State-based Steering Stéphane Caron a,c, Quang-Cuong Pham b, Yoshihiko Nakamura a arxiv:1511.05259v2 [cs.ro] 20 Nov 2015 a Department of Mechano-Informatics,

More information

Classification: The rest of the story

Classification: The rest of the story U NIVERSITY OF ILLINOIS AT URBANA-CHAMPAIGN CS598 Machine Learning for Signal Processing Classification: The rest of the story 3 October 2017 Today s lecture Important things we haven t covered yet Fisher

More information

The Nonlinear Velocity Obstacle Revisited: the Optimal Time Horizon

The Nonlinear Velocity Obstacle Revisited: the Optimal Time Horizon The Nonlinear Velocity Obstacle Revisited: the Optimal Time Horizon Zvi Shiller, Oren Gal, and Thierry Fraichard bstract This paper addresses the issue of motion safety in dynamic environments using Velocity

More information

Improving the Performance of Sampling-Based Planners by Using a. Symmetry-Exploiting Gap Reduction Algorithm

Improving the Performance of Sampling-Based Planners by Using a. Symmetry-Exploiting Gap Reduction Algorithm Improving the Performance of Sampling-Based Planners by Using a Symmetry-Exploiting Gap Reduction Algorithm Peng Cheng Emilio Frazzoli Steven M. LaValle University of Illinois Urbana, IL 61801 USA {pcheng1,

More information

Sufficient Conditions for the Existence of Resolution Complete Planning Algorithms

Sufficient Conditions for the Existence of Resolution Complete Planning Algorithms Sufficient Conditions for the Existence of Resolution Complete Planning Algorithms Dmitry Yershov and Steve LaValle Computer Science niversity of Illinois at rbana-champaign WAFR 2010 December 15, 2010

More information

Approximate Voronoi Diagrams

Approximate Voronoi Diagrams CS468, Mon. Oct. 30 th, 2006 Approximate Voronoi Diagrams Presentation by Maks Ovsjanikov S. Har-Peled s notes, Chapters 6 and 7 1-1 Outline Preliminaries Problem Statement ANN using PLEB } Bounds and

More information

Potential Field Methods

Potential Field Methods Randomized Motion Planning Nancy Amato Fall 04, Univ. of Padova Potential Field Methods [ ] Potential Field Methods Acknowledgement: Parts of these course notes are based on notes from courses given by

More information

A Multi-Directional Rapidly Exploring Random Graph (mrrg) for Protein Folding

A Multi-Directional Rapidly Exploring Random Graph (mrrg) for Protein Folding A Multi-Directional Rapidly Exploring Random Graph (mrrg) for Protein Folding Shuvra Kanti Nath, Shawna Thomas, Chinwe Ekenna, and Nancy M. Amato Parasol Lab, Department of Computer Science and Engineering

More information

Exploiting Context and Semantics for UAV Path-finding in an Urban Setting

Exploiting Context and Semantics for UAV Path-finding in an Urban Setting Exploiting Context and Semantics for UAV Path-finding in an Urban Setting Marjan Alirezaie, Andrey Kiselev, Franziska Klügl, Martin Längkvist, and Amy Loutfi Machine Perception and Interaction Lab, AASS

More information

Motion Planning in Partially Known Dynamic Environments

Motion Planning in Partially Known Dynamic Environments Motion Planning in Partially Known Dynamic Environments Movie Workshop Laas-CNRS, Toulouse (FR), January 7-8, 2005 Dr. Thierry Fraichard e-motion Team Inria Rhône-Alpes & Gravir-CNRS Laboratory Movie Workshop

More information

A Randomized Tree Construction Algorithm to Explore Energy Landscapes

A Randomized Tree Construction Algorithm to Explore Energy Landscapes A Randomized Tree Construction Algorithm to Explore Energy Landscapes L. Jaillet 1, F.J. Corcho 2, J.J. Pérez 2, J. Cortés 3,4 1 InstitutdeRobòticaiInformàticaIndustrial, CSIC-UPC,C/LlorensiArtigas4-6,

More information

An Adaptive Multi-resolution State Lattice Approach for Motion Planning with Uncertainty

An Adaptive Multi-resolution State Lattice Approach for Motion Planning with Uncertainty An Adaptive Multi-resolution State Lattice Approach for Motion Planning with Uncertainty A. González-Sieira 1, M. Mucientes 1 and A. Bugarín 1 Centro de Investigación en Tecnoloxías da Información (CiTIUS),

More information

Gross Motion Planning

Gross Motion Planning Gross Motion Planning...given a moving object, A, initially in an unoccupied region of freespace, s, a set of stationary objects, B i, at known locations, and a legal goal position, g, find a sequence

More information

QUANTIZED SYSTEMS AND CONTROL. Daniel Liberzon. DISC HS, June Dept. of Electrical & Computer Eng., Univ. of Illinois at Urbana-Champaign

QUANTIZED SYSTEMS AND CONTROL. Daniel Liberzon. DISC HS, June Dept. of Electrical & Computer Eng., Univ. of Illinois at Urbana-Champaign QUANTIZED SYSTEMS AND CONTROL Daniel Liberzon Coordinated Science Laboratory and Dept. of Electrical & Computer Eng., Univ. of Illinois at Urbana-Champaign DISC HS, June 2003 HYBRID CONTROL Plant: u y

More information

10 Robotic Exploration and Information Gathering

10 Robotic Exploration and Information Gathering NAVARCH/EECS 568, ROB 530 - Winter 2018 10 Robotic Exploration and Information Gathering Maani Ghaffari April 2, 2018 Robotic Information Gathering: Exploration and Monitoring In information gathering

More information

ECE 307 Techniques for Engineering Decisions

ECE 307 Techniques for Engineering Decisions ECE 7 Techniques for Engineering Decisions Introduction to the Simple Algorithm George Gross Department of Electrical and Computer Engineering University of Illinois at Urbana-Champaign ECE 7 5 9 George

More information

Planning Persistent Monitoring Trajectories in Spatiotemporal Fields using Kinodynamic RRT*

Planning Persistent Monitoring Trajectories in Spatiotemporal Fields using Kinodynamic RRT* Planning Persistent Monitoring Trajectories in Spatiotemporal Fields using Kinodynamic RRT* Xiaodong Lan, Cristian-Ioan Vasile, and Mac Schwager Abstract This paper proposes a sampling based kinodynamic

More information

Protein Folding by Robotics

Protein Folding by Robotics Protein Folding by Robotics 1 TBI Winterseminar 2006 February 21, 2006 Protein Folding by Robotics 1 TBI Winterseminar 2006 February 21, 2006 Protein Folding by Robotics Probabilistic Roadmap Planning

More information

Laplacian-Centered Poisson Solvers and Multilevel Summation Algorithms

Laplacian-Centered Poisson Solvers and Multilevel Summation Algorithms Laplacian-Centered Poisson Solvers and Multilevel Summation Algorithms Dmitry Yershov 1 Stephen Bond 1 Robert Skeel 2 1 University of Illinois at Urbana-Champaign 2 Purdue University 2009 SIAM Annual Meeting

More information

CS 473: Algorithms. Ruta Mehta. Spring University of Illinois, Urbana-Champaign. Ruta (UIUC) CS473 1 Spring / 32

CS 473: Algorithms. Ruta Mehta. Spring University of Illinois, Urbana-Champaign. Ruta (UIUC) CS473 1 Spring / 32 CS 473: Algorithms Ruta Mehta University of Illinois, Urbana-Champaign Spring 2018 Ruta (UIUC) CS473 1 Spring 2018 1 / 32 CS 473: Algorithms, Spring 2018 Universal Hashing Lecture 10 Feb 15, 2018 Most

More information

Algorithms for Picture Analysis. Lecture 07: Metrics. Axioms of a Metric

Algorithms for Picture Analysis. Lecture 07: Metrics. Axioms of a Metric Axioms of a Metric Picture analysis always assumes that pictures are defined in coordinates, and we apply the Euclidean metric as the golden standard for distance (or derived, such as area) measurements.

More information

Robust Adaptive Motion Planning in the Presence of Dynamic Obstacles

Robust Adaptive Motion Planning in the Presence of Dynamic Obstacles 2016 American Control Conference (ACC) Boston Marriott Copley Place July 6-8, 2016. Boston, MA, USA Robust Adaptive Motion Planning in the Presence of Dynamic Obstacles Nurali Virani and Minghui Zhu Abstract

More information

Towards Reduced-Order Models for Online Motion Planning and Control of UAVs in the Presence of Wind

Towards Reduced-Order Models for Online Motion Planning and Control of UAVs in the Presence of Wind Towards Reduced-Order Models for Online Motion Planning and Control of UAVs in the Presence of Wind Ashray A. Doshi, Surya P. Singh and Adam J. Postula The University of Queensland, Australia {a.doshi,

More information

Improving the Performance of Sampling-Based Motion Planning With Symmetry-Based Gap Reduction

Improving the Performance of Sampling-Based Motion Planning With Symmetry-Based Gap Reduction 488 IEEE TRANSACTIONS ON ROBOTICS, VOL. 24, NO. 2, APRIL 2008 Improving the Performance of Sampling-Based Motion Planning With Symmetry-Based Gap Reduction Peng Cheng, Emilio Frazzoli, and Steven LaValle

More information

Beyond Basic Path Planning in C-Space

Beyond Basic Path Planning in C-Space Beyond Basic Path Planning in C-Space Steve LaValle University of Illinois September 30, 2011 IROS 2011 Workshop Presentation September 2011 1 / 31 Three Main Places for Prospecting After putting together

More information

A Topologically Constrained Ising Model

A Topologically Constrained Ising Model Constrained Ising Model July 29, 2016 1 / 50 A Topologically Constrained Ising Model Derek Kielty UIUC Katherine Koch UIUC William Linz UIUC Colleen Robichaux UIUC Fernando Roman-Garcia UIUC Elizabeth

More information

SESSION 2 MULTI-AGENT NETWORKS. Magnus Egerstedt - Aug. 2013

SESSION 2 MULTI-AGENT NETWORKS. Magnus Egerstedt - Aug. 2013 SESSION 2 MULTI-AGENT NETWORKS Variations on the Theme: Directed Graphs Instead of connectivity, we need directed notions: Strong connectivity = there exists a directed path between any two nodes Weak

More information

Hideyuki Usui 1,3, M. Nunami 2,3, Y. Yagi 1,3, T. Moritaka 1,3, and JST/CREST multi-scale PIC simulation team

Hideyuki Usui 1,3, M. Nunami 2,3, Y. Yagi 1,3, T. Moritaka 1,3, and JST/CREST multi-scale PIC simulation team Hideyuki Usui 1,3, M. Nunami 2,3, Y. Yagi 1,3, T. Moritaka 1,3, and JST/CREST multi-scale PIC simulation team 1 Kobe Univ., Japan, 2 NIFS,Japan, 3 JST/CREST, Outline Multi-scale interaction between weak

More information

Optimal Data-Dependent Hashing for Approximate Near Neighbors

Optimal Data-Dependent Hashing for Approximate Near Neighbors Optimal Data-Dependent Hashing for Approximate Near Neighbors Alexandr Andoni 1 Ilya Razenshteyn 2 1 Simons Institute 2 MIT, CSAIL April 20, 2015 1 / 30 Nearest Neighbor Search (NNS) Let P be an n-point

More information

Force Fields for Classical Molecular Dynamics simulations of Biomolecules. Emad Tajkhorshid

Force Fields for Classical Molecular Dynamics simulations of Biomolecules. Emad Tajkhorshid Force Fields for Classical Molecular Dynamics simulations of Biomolecules Emad Tajkhorshid Theoretical and Computational Biophysics Group, Beckman Institute Departments of Biochemistry and Pharmacology,

More information

Bounds on Tracking Error using Closed-Loop Rapidly-Exploring Random Trees

Bounds on Tracking Error using Closed-Loop Rapidly-Exploring Random Trees Bounds on Tracking Error using Closed-Loop Rapidly-Eploring Random Trees Brandon D. Luders, Sertac Karaman, Emilio Frazzoli, and Jonathan P. How Abstract This paper considers the real-time motion planning

More information

Realtime Informed Path Sampling for Motion Planning Search

Realtime Informed Path Sampling for Motion Planning Search Realtime Informed Path Sampling for Motion Planning Search Ross A. Knepper and Matthew T. Mason Abstract Robot motions typically originate from an uninformed path sampling process such as random or low-dispersion

More information

DryVR: Data-driven verification and compositional reasoning for automotive systems

DryVR: Data-driven verification and compositional reasoning for automotive systems DryVR: Data-driven verification and compositional reasoning for automotive systems Chuchu Fan, Bolun Qi, Sayan Mitra, Mahesh Viswannathan University of Illinois at Urbana-Champaign CAV 2017, Heidelberg,

More information

Holdout and Cross-Validation Methods Overfitting Avoidance

Holdout and Cross-Validation Methods Overfitting Avoidance Holdout and Cross-Validation Methods Overfitting Avoidance Decision Trees Reduce error pruning Cost-complexity pruning Neural Networks Early stopping Adjusting Regularizers via Cross-Validation Nearest

More information

Path Planning using Positive Invariant Sets

Path Planning using Positive Invariant Sets MITSUBISHI ELECTRIC RESEARCH LABORATORIES http://www.merl.com Path Planning using Positive Invariant Sets Danielson, C.; Weiss, A.; Berntorp, K.; Di Cairano, S. TR2017-046 March 2017 Abstract We present

More information

Geometric View of Machine Learning Nearest Neighbor Classification. Slides adapted from Prof. Carpuat

Geometric View of Machine Learning Nearest Neighbor Classification. Slides adapted from Prof. Carpuat Geometric View of Machine Learning Nearest Neighbor Classification Slides adapted from Prof. Carpuat What we know so far Decision Trees What is a decision tree, and how to induce it from data Fundamental

More information

CMSC 422 Introduction to Machine Learning Lecture 4 Geometry and Nearest Neighbors. Furong Huang /

CMSC 422 Introduction to Machine Learning Lecture 4 Geometry and Nearest Neighbors. Furong Huang / CMSC 422 Introduction to Machine Learning Lecture 4 Geometry and Nearest Neighbors Furong Huang / furongh@cs.umd.edu What we know so far Decision Trees What is a decision tree, and how to induce it from

More information

Lecture 14: Multiple Sequence Alignment (Gene Finding, Conserved Elements) Scribe: John Ekins

Lecture 14: Multiple Sequence Alignment (Gene Finding, Conserved Elements) Scribe: John Ekins Lecture 14: Multiple Sequence Alignment (Gene Finding, Conserved Elements) 2 19 2015 Scribe: John Ekins Multiple Sequence Alignment Given N sequences x 1, x 2,, x N : Insert gaps in each of the sequences

More information

The Lattice Boltzmann Method for Laminar and Turbulent Channel Flows

The Lattice Boltzmann Method for Laminar and Turbulent Channel Flows The Lattice Boltzmann Method for Laminar and Turbulent Channel Flows Vanja Zecevic, Michael Kirkpatrick and Steven Armfield Department of Aerospace Mechanical & Mechatronic Engineering The University of

More information

Transforming Hierarchical Trees on Metric Spaces

Transforming Hierarchical Trees on Metric Spaces CCCG 016, Vancouver, British Columbia, August 3 5, 016 Transforming Hierarchical Trees on Metric Spaces Mahmoodreza Jahanseir Donald R. Sheehy Abstract We show how a simple hierarchical tree called a cover

More information

Controllable kinematic reductions for mechanical systems: concepts, computational tools, and examples

Controllable kinematic reductions for mechanical systems: concepts, computational tools, and examples Proceedings of MTNS02 University of Notre Dame, August 2002 2 F. Bullo and A. D. Lewis and K. M. Lynch Controllable kinematic reductions for mechanical systems: concepts, computational tools, and examples

More information

Asymptotically Stable Gait Primitives for Planning Dynamic Bipedal Locomotion in Three Dimensions

Asymptotically Stable Gait Primitives for Planning Dynamic Bipedal Locomotion in Three Dimensions May 4, 2010 ICRA, Anchorage, AK 1 1 Asymptotically Stable Gait Primitives for Planning Dynamic Bipedal Locomotion in Three Dimensions Robert D. Gregg *, Timothy W. Bretl, Mark W. Spong Coordinated Science

More information

Kinodynamic Randomized Rearrangement Planning via Dynamic Transitions Between Statically Stable States

Kinodynamic Randomized Rearrangement Planning via Dynamic Transitions Between Statically Stable States 2015 IEEE International Conference on Robotics and Automation (ICRA) Washington State Convention Center Seattle, Washington, May 26-30, 2015 Kinodynamic Randomized Rearrangement Planning via Dynamic Transitions

More information

2 If ax + bx + c = 0, then x = b) What are the x-intercepts of the graph or the real roots of f(x)? Round to 4 decimal places.

2 If ax + bx + c = 0, then x = b) What are the x-intercepts of the graph or the real roots of f(x)? Round to 4 decimal places. Quadratic Formula - Key Background: So far in this course we have solved quadratic equations by the square root method and the factoring method. Each of these methods has its strengths and limitations.

More information

Sampling-based path planning: a new tool for missile guidance

Sampling-based path planning: a new tool for missile guidance Sampling-based path planning: a new tool for missile guidance P. Pharpatara, B. Hérissé, R. Pepy, Y. Bestaoui To cite this version: P. Pharpatara, B. Hérissé, R. Pepy, Y. Bestaoui. Sampling-based path

More information

CSE 417T: Introduction to Machine Learning. Final Review. Henry Chai 12/4/18

CSE 417T: Introduction to Machine Learning. Final Review. Henry Chai 12/4/18 CSE 417T: Introduction to Machine Learning Final Review Henry Chai 12/4/18 Overfitting Overfitting is fitting the training data more than is warranted Fitting noise rather than signal 2 Estimating! "#$

More information

Adapting Boyer-Moore-Like Algorithms for Searching Huffman Encoded Texts

Adapting Boyer-Moore-Like Algorithms for Searching Huffman Encoded Texts Adapting Boyer-Moore-Like Algorithms for Searching Huffman Encoded Texts Domenico Cantone Simone Faro Emanuele Giaquinta Department of Mathematics and Computer Science, University of Catania, Italy 1 /

More information

Lecture 18: March 15

Lecture 18: March 15 CS71 Randomness & Computation Spring 018 Instructor: Alistair Sinclair Lecture 18: March 15 Disclaimer: These notes have not been subjected to the usual scrutiny accorded to formal publications. They may

More information

12. LOCAL SEARCH. gradient descent Metropolis algorithm Hopfield neural networks maximum cut Nash equilibria

12. LOCAL SEARCH. gradient descent Metropolis algorithm Hopfield neural networks maximum cut Nash equilibria Coping With NP-hardness Q. Suppose I need to solve an NP-hard problem. What should I do? A. Theory says you re unlikely to find poly-time algorithm. Must sacrifice one of three desired features. Solve

More information

Clustering Perturbation Resilient

Clustering Perturbation Resilient Clustering Perturbation Resilient Instances Maria-Florina Balcan Carnegie Mellon University Clustering Comes Up Everywhere Clustering news articles or web pages or search results by topic. Clustering protein

More information

Planning as Satisfiability

Planning as Satisfiability Planning as Satisfiability Alan Fern * Review of propositional logic (see chapter 7) Planning as propositional satisfiability Satisfiability techniques (see chapter 7) Combining satisfiability techniques

More information

Planning Periodic Persistent Monitoring Trajectories for Sensing Robots in Gaussian Random Fields

Planning Periodic Persistent Monitoring Trajectories for Sensing Robots in Gaussian Random Fields 013 IEEE International Conference on Robotics and Automation (ICRA) Karlsruhe, Germany, May 6-10, 013 Planning Periodic Persistent Monitoring Trajectories for Sensing Robots in Gaussian Random Fields Xiaodong

More information

Delta-net: Real-time Network Verification Using Atoms

Delta-net: Real-time Network Verification Using Atoms Delta-net: Real-time Network Verification Using Atoms Alex Horn 1, Ali Kheradmand 2 and Mukul R. Prasad 1 1 Fujitsu Labs of America 2 University of Illinois at Urbana-Champaign (Internship) NSDI 2017,

More information

Probabilistic Feasibility for Nonlinear Systems with Non-Gaussian Uncertainty using RRT

Probabilistic Feasibility for Nonlinear Systems with Non-Gaussian Uncertainty using RRT Probabilistic Feasibility for Nonlinear Systems with Non-Gaussian Uncertainty using RRT Brandon Luders and Jonathan P. How Aerospace Controls Laboratory Massachusetts Institute of Technology, Cambridge,

More information

Composable Group Behaviors

Composable Group Behaviors Composable Group Behaviors Perpetual Amoah Undergraduate S. Researcher Sam Rodriguez Graduate Student Mentor Jyh-Ming Lien Graduate Student Mentor John Maffei Graduate Student Mentor Nancy M. Amato Faculty

More information

From CDF to PDF A Density Estimation Method for High Dimensional Data

From CDF to PDF A Density Estimation Method for High Dimensional Data From CDF to PDF A Density Estimation Method for High Dimensional Data Shengdong Zhang Simon Fraser University sza75@sfu.ca arxiv:1804.05316v1 [stat.ml] 15 Apr 2018 April 17, 2018 1 Introduction Probability

More information

Sampling-based motion planning with deterministic u- calculus specifications

Sampling-based motion planning with deterministic u- calculus specifications Sampling-based motion planning with deterministic u- calculus specifications The MIT Faculty has made this article openly available. Please share how this access benefits you. Your story matters. Citation

More information

Integration of Local Geometry and Metric Information in Sampling-Based Motion Planning

Integration of Local Geometry and Metric Information in Sampling-Based Motion Planning University of Pennsylvania ScholarlyCommons Departmental Papers (ESE) Department of Electrical & Systems Engineering 2-25-2018 Integration of Local Geometry and Metric Information in Sampling-Based Motion

More information

Application of Ultrasonic Wave Reflection for Setting and Stiffening of Cement Paste

Application of Ultrasonic Wave Reflection for Setting and Stiffening of Cement Paste Application of Ultrasonic Wave Reflection for Setting and Stiffening of Cement Paste *Chul-Woo Chung 1) John S. Popovics 2) and Leslie J. Struble 3) 1) Department of Architectural Engineering, Pukyong

More information

9/26/17. Ridge regression. What our model needs to do. Ridge Regression: L2 penalty. Ridge coefficients. Ridge coefficients

9/26/17. Ridge regression. What our model needs to do. Ridge Regression: L2 penalty. Ridge coefficients. Ridge coefficients What our model needs to do regression Usually, we are not just trying to explain observed data We want to uncover meaningful trends And predict future observations Our questions then are Is β" a good estimate

More information

Algorithms for Calculating Statistical Properties on Moving Points

Algorithms for Calculating Statistical Properties on Moving Points Algorithms for Calculating Statistical Properties on Moving Points Dissertation Proposal Sorelle Friedler Committee: David Mount (Chair), William Gasarch Samir Khuller, Amitabh Varshney January 14, 2009

More information

ICTP Conference Graphene Week 2008

ICTP Conference Graphene Week 2008 1960-3 ICTP Conference Graphene Week 2008 25-29 August 2008 Current-induced cleaning of graphene J. Moser CIN2 Barcelona, Campus UAB, Bellaterra, Spain A. Barreiro CIN2 Barcelona, Campus UAB, Bellaterra,

More information

On the Cost of Worst-Case Coding Length Constraints

On the Cost of Worst-Case Coding Length Constraints On the Cost of Worst-Case Coding Length Constraints Dror Baron and Andrew C. Singer Abstract We investigate the redundancy that arises from adding a worst-case length-constraint to uniquely decodable fixed

More information

Incremental Proof-Based Verification of Compiler Optimizations

Incremental Proof-Based Verification of Compiler Optimizations Incremental Proof-Based Verification of Compiler Optimizations Grigory Fedyukovich joint work with Arie Gurfinkel and Natasha Sharygina 5 of May, 2015, Attersee, Austria counter-example change impact Big

More information

CS 781 Lecture 9 March 10, 2011 Topics: Local Search and Optimization Metropolis Algorithm Greedy Optimization Hopfield Networks Max Cut Problem Nash

CS 781 Lecture 9 March 10, 2011 Topics: Local Search and Optimization Metropolis Algorithm Greedy Optimization Hopfield Networks Max Cut Problem Nash CS 781 Lecture 9 March 10, 2011 Topics: Local Search and Optimization Metropolis Algorithm Greedy Optimization Hopfield Networks Max Cut Problem Nash Equilibrium Price of Stability Coping With NP-Hardness

More information

RRT X : Real-Time Motion Planning/Replanning for Environments with Unpredictable Obstacles

RRT X : Real-Time Motion Planning/Replanning for Environments with Unpredictable Obstacles RRT X : Real-Time Motion Planning/Replanning for Environments with Unpredictable Obstacles Michael Otte and Emilio Frazzoli Massachusetts Institute of Technology, Cambridge MA 2139, USA ottemw@mit.edu

More information

4 Locality-sensitive hashing using stable distributions

4 Locality-sensitive hashing using stable distributions 4 Locality-sensitive hashing using stable distributions 4. The LSH scheme based on s-stable distributions In this chapter, we introduce and analyze a novel locality-sensitive hashing family. The family

More information

A Fast and Simple Algorithm for Computing Approximate Euclidean Minimum Spanning Trees

A Fast and Simple Algorithm for Computing Approximate Euclidean Minimum Spanning Trees A Fast and Simple Algorithm for Computing Approximate Euclidean Minimum Spanning Trees Sunil Arya Hong Kong University of Science and Technology and David Mount University of Maryland Arya and Mount HALG

More information

CS 273 Prof. Serafim Batzoglou Prof. Jean-Claude Latombe Spring Lecture 12 : Energy maintenance (1) Lecturer: Prof. J.C.

CS 273 Prof. Serafim Batzoglou Prof. Jean-Claude Latombe Spring Lecture 12 : Energy maintenance (1) Lecturer: Prof. J.C. CS 273 Prof. Serafim Batzoglou Prof. Jean-Claude Latombe Spring 2006 Lecture 12 : Energy maintenance (1) Lecturer: Prof. J.C. Latombe Scribe: Neda Nategh How do you update the energy function during the

More information

14 : Approximate Inference Monte Carlo Methods

14 : Approximate Inference Monte Carlo Methods 10-708: Probabilistic Graphical Models 10-708, Spring 2018 14 : Approximate Inference Monte Carlo Methods Lecturer: Kayhan Batmanghelich Scribes: Biswajit Paria, Prerna Chiersal 1 Introduction We have

More information

neutrinos (ν) } ν energy ~ K ν + proton e + + neutron! e - + proton neutron + ν Freeze-out temperatures

neutrinos (ν) } ν energy ~ K ν + proton e + + neutron! e - + proton neutron + ν Freeze-out temperatures kt ~ mparticle c 2 neutrinos (ν) kt < mparticle c 2 kt > mparticle c 2 Freeze-out temperatures particle /! T (K) time since BB antiparticle 6x10 e 20 sec 1.2x10 μ 1 sec 1x10 p 10 NOTE: after freeze-out,

More information

Robust Low Torque Biped Walking Using Differential Dynamic Programming With a Minimax Criterion

Robust Low Torque Biped Walking Using Differential Dynamic Programming With a Minimax Criterion Robust Low Torque Biped Walking Using Differential Dynamic Programming With a Minimax Criterion J. Morimoto and C. Atkeson Human Information Science Laboratories, ATR International, Department 3 2-2-2

More information

Algorithms for Data Science: Lecture on Finding Similar Items

Algorithms for Data Science: Lecture on Finding Similar Items Algorithms for Data Science: Lecture on Finding Similar Items Barna Saha 1 Finding Similar Items Finding similar items is a fundamental data mining task. We may want to find whether two documents are similar

More information

A Decision Stump. Decision Trees, cont. Boosting. Machine Learning 10701/15781 Carlos Guestrin Carnegie Mellon University. October 1 st, 2007

A Decision Stump. Decision Trees, cont. Boosting. Machine Learning 10701/15781 Carlos Guestrin Carnegie Mellon University. October 1 st, 2007 Decision Trees, cont. Boosting Machine Learning 10701/15781 Carlos Guestrin Carnegie Mellon University October 1 st, 2007 1 A Decision Stump 2 1 The final tree 3 Basic Decision Tree Building Summarized

More information

Metrics: Growth, dimension, expansion

Metrics: Growth, dimension, expansion Metrics: Growth, dimension, expansion Social and Technological Networks Rik Sarkar University of Edinburgh, 2017. Metric A distance measure d is a metric if: d(u,v) 0 d(u,v) = 0 iff u=v d(u,v) = d(u,v)

More information

Wolf-Tilo Balke Silviu Homoceanu Institut für Informationssysteme Technische Universität Braunschweig

Wolf-Tilo Balke Silviu Homoceanu Institut für Informationssysteme Technische Universität Braunschweig Multimedia Databases Wolf-Tilo Balke Silviu Homoceanu Institut für Informationssysteme Technische Universität Braunschweig http://www.ifis.cs.tu-bs.de 13 Indexes for Multimedia Data 13 Indexes for Multimedia

More information

Space-Time Tradeoffs for Approximate Spherical Range Counting

Space-Time Tradeoffs for Approximate Spherical Range Counting Space-Time Tradeoffs for Approximate Spherical Range Counting Sunil Arya Theocharis Malamatos David M. Mount University of Maryland Technical Report CS TR 4842 and UMIACS TR 2006 57 November 2006 Abstract

More information

COMP3702/7702 Artificial Intelligence Week 5: Search in Continuous Space with an Application in Motion Planning " Hanna Kurniawati"

COMP3702/7702 Artificial Intelligence Week 5: Search in Continuous Space with an Application in Motion Planning  Hanna Kurniawati COMP3702/7702 Artificial Intelligence Week 5: Search in Continuous Space with an Application in Motion Planning " Hanna Kurniawati" Last week" Main components of PRM" Collision check for a configuration"

More information

Building a Multi-FPGA Virtualized Restricted Boltzmann Machine Architecture Using Embedded MPI

Building a Multi-FPGA Virtualized Restricted Boltzmann Machine Architecture Using Embedded MPI Building a Multi-FPGA Virtualized Restricted Boltzmann Machine Architecture Using Embedded MPI Charles Lo and Paul Chow {locharl1, pc}@eecg.toronto.edu Department of Electrical and Computer Engineering

More information

On and Off-Policy Relational Reinforcement Learning

On and Off-Policy Relational Reinforcement Learning On and Off-Policy Relational Reinforcement Learning Christophe Rodrigues, Pierre Gérard, and Céline Rouveirol LIPN, UMR CNRS 73, Institut Galilée - Université Paris-Nord first.last@lipn.univ-paris13.fr

More information

Br OAc. OAc. Problem R-11L (C 16 H 21 BrO 10 ) 270 MHz 1 H NMR spectrum in CDCl 3 Source: Ieva Reich (digitized hard copy) g. AcO. AcO H.

Br OAc. OAc. Problem R-11L (C 16 H 21 BrO 10 ) 270 MHz 1 H NMR spectrum in CDCl 3 Source: Ieva Reich (digitized hard copy) g. AcO. AcO H. Problem R-11L (C 16 21 O ) 270 Mz 1 NMR spectrum in CDCl 3 30 20 0 z 1541.1 1544.3 1530.9 1534.3 1501.8 1506.0 1509.4 1511.3 1512.6 1492.1 1476.0 1466.4 1456.8 1403.1 1406.8 1393.2 1397.2 1198.2 1201.9

More information

Controllable kinematic reductions for mechanical systems: concepts, computational tools, and examples

Controllable kinematic reductions for mechanical systems: concepts, computational tools, and examples Controllable kinematic reductions for mechanical systems: concepts, computational tools, and examples Francesco Bullo Coordinated Science Lab University of Illinois Urbana-Champaign Urbana, IL 61801, USA

More information

Bandit View on Continuous Stochastic Optimization

Bandit View on Continuous Stochastic Optimization Bandit View on Continuous Stochastic Optimization Sébastien Bubeck 1 joint work with Rémi Munos 1 & Gilles Stoltz 2 & Csaba Szepesvari 3 1 INRIA Lille, SequeL team 2 CNRS/ENS/HEC 3 University of Alberta

More information

A Motion Planning Approach to Folding:

A Motion Planning Approach to Folding: A Motion Planning Approach to Folding: From Paper Craft to Protein Folding Λ Guang Song Nancy M. Amato Department of Computer Science Texas A&M University College Station, TX 77843-32 fgsong,amatog@cs.tamu.edu

More information

Ridge Regression: Regulating overfitting when using many features. Training, true, & test error vs. model complexity. CSE 446: Machine Learning

Ridge Regression: Regulating overfitting when using many features. Training, true, & test error vs. model complexity. CSE 446: Machine Learning Ridge Regression: Regulating overfitting when using many features Emily Fox University of Washington January 3, 207 Training, true, & test error vs. model complexity Overfitting if: Error y Model complexity

More information

xkcd.com It IS about physics. It ALL is.

xkcd.com It IS about physics. It ALL is. xkcd.com It IS about physics. It ALL is. Introduction to Space Plasmas The Plasma State What is a plasma? Basic plasma properties: Qualitative & Quantitative Examples of plasmas Single particle motion

More information