Muutke küpsiste eelistusi

Data Mining for Social Robotics: Toward Autonomously Social Robots 1st ed. 2015 [Kõva köide]

4.33/5 (6 hinnangut Goodreads-ist)
,
  • Formaat: Hardback, 328 pages, kõrgus x laius: 235x155 mm, kaal: 6328 g, 74 Illustrations, color; XII, 328 p. 74 illus. in color., 1 Hardback
  • Sari: Advanced Information and Knowledge Processing
  • Ilmumisaeg: 10-Feb-2016
  • Kirjastus: Springer International Publishing AG
  • ISBN-10: 3319252305
  • ISBN-13: 9783319252308
Teised raamatud teemal:
  • Kõva köide
  • Hind: 95,02 €*
  • * hind on lõplik, st. muud allahindlused enam ei rakendu
  • Tavahind: 111,79 €
  • Säästad 15%
  • Raamatu kohalejõudmiseks kirjastusest kulub orienteeruvalt 2-4 nädalat
  • Kogus:
  • Lisa ostukorvi
  • Tasuta tarne
  • Tellimisaeg 2-4 nädalat
  • Lisa soovinimekirja
Data Mining for Social Robotics: Toward Autonomously Social Robots 1st ed. 2015Suurem pilt
  • Formaat: Hardback, 328 pages, kõrgus x laius: 235x155 mm, kaal: 6328 g, 74 Illustrations, color; XII, 328 p. 74 illus. in color., 1 Hardback
  • Sari: Advanced Information and Knowledge Processing
  • Ilmumisaeg: 10-Feb-2016
  • Kirjastus: Springer International Publishing AG
  • ISBN-10: 3319252305
  • ISBN-13: 9783319252308
Teised raamatud teemal:
Püsilink: https://www.kriso.ee/db/9783319252308.html
Märksõnad:
This book explores an approach to social robotics based solely on autonomous unsupervised techniques and positions it within a structured exposition of related research in psychology, neuroscience, HRI, and data mining.  The authors present an autonomous and developmental approach that allows the robot to learn interactive behavior by imitating humans using algorithms from time-series analysis and machine learning.

The first part provides a comprehensive and structured introduction to time-series analysis, change point discovery, motif discovery and causality analysis focusing on possible applicability to HRI problems. Detailed explanations of all the algorithms involved are provided with open-source implementations in MATLAB enabling the reader to experiment with them. Imitation and simulation are the key technologies used to attain social behavior autonomously in the proposed approach.  Part two gives the reader a wide overview of research in these areas in psychology, and ethology. Based on this background, the authors discuss approaches to endow robots with the ability to autonomously learn how to be social. 





Data Mining for Social Robots will be essential reading for graduate students and practitioners interested in social and developmental robotics.

Arvustused

This comprehensive work focuses on human-robot interaction (HRI) using data mining and time series analysis. In general, this book includes rich knowledge in social robot study using data mining tools. Its a nice book for graduate students and practitioners to dive deeper into HRI. Personally, this book led me to rethink the learning processes and interaction manners of humans, which is a rather interesting journey. (Feng Yu, Computing Reviews, March, 2017)

1 Introduction
1(34)
1.1 Motivation
1(4)
1.2 General Overview
5(2)
1.3 Relation to Different Research Fields
7(5)
1.3.1 Interaction Studies
7(2)
1.3.2 Robotics
9(2)
1.3.3 Neuroscience and Experimental Psychology
11(1)
1.3.4 Machine Learning and Data Mining
11(1)
1.3.5 Contributions
11(1)
1.4 Interaction Scenarios
12(2)
1.5 Nonverbal Communication in Human-Human Interactions
14(3)
1.6 Nonverbal Communication in Human-Robot Interactions
17(4)
1.6.1 Appearance
18(1)
1.6.2 Gesture Interfaces
18(1)
1.6.3 Spontaneous Nonverbal Behavior
19(2)
1.7 Behavioral Robotic Architectures
21(3)
1.7.1 Reactive Architectures
21(1)
1.7.2 Hybrid Architectures
22(1)
1.7.3 HRI Specific Architectures
23(1)
1.8 Learning from Demonstrations
24(2)
1.9 Book Organization
26(1)
1.10 Supporting Site
27(1)
1.11 Summary
28(7)
References
28(7)
Part I Time Series Mining
2 Mining Time-Series Data
35(50)
2.1 Basic Definitions
35(1)
2.2 Models of Time-Series Generating Processes
36(14)
2.2.1 Linear Additive Time-Series Model
36(1)
2.2.2 Random Walk
37(1)
2.2.3 Moving Average Processes
38(2)
2.2.4 Auto-Regressive Processes
40(1)
2.2.5 ARMA and ARIMA Processes
40(1)
2.2.6 State-Space Generation
41(1)
2.2.7 Markov Chains
42(1)
2.2.8 Hidden Markov Models
43(2)
2.2.9 Gaussian Mixture Models
45(2)
2.2.10 Gaussian Processes
47(3)
2.3 Representation and Transformations
50(17)
2.3.1 Piecewise Aggregate Approximation
51(1)
2.3.2 Symbolic Aggregate Approximation
52(2)
2.3.3 Discrete Fourier Transform
54(1)
2.3.4 Discrete Wavelet Transform
55(1)
2.3.5 Singular Spectrum Analysis
56(11)
2.4 Learning Time-Series Models from Data
67(10)
2.4.1 Learning an AR Process
67(3)
2.4.2 Learning an ARMA Process
70(3)
2.4.3 Learning a Hidden Markov Model
73(3)
2.4.4 Learning a Gaussian Mixture Model
76(1)
2.4.5 Model Selection Problem
77(1)
2.5 Time Series Preprocessing
77(5)
2.5.1 Smoothing
77(1)
2.5.2 Thinning
78(1)
2.5.3 Normalization
78(1)
2.5.4 De-Trending
79(1)
2.5.5 Dimensionality Reduction
80(1)
2.5.6 Dynamic Time Warping
81(1)
2.6 Summary
82(3)
References
83(2)
3 Change Point Discovery
85(24)
3.1 Approaches to CP Discovery
86(1)
3.2 Markov Process CP Approach
87(3)
3.3 Two Models Approach
90(3)
3.4 Change in Stochastic Processes
93(1)
3.5 Singular Spectrum Analysis Based Methods
94(4)
3.5.1 Alternative SSA CPD Methods
98(1)
3.6 Change Localization
98(1)
3.7 Comparing CPD Algorithms
99(6)
3.7.1 Confusion Matrix Measures
100(1)
3.7.2 Divergence Measures
101(3)
3.7.3 Equal Sampling Rate
104(1)
3.8 CPD for Measuring Naturalness in HRI
105(2)
3.9 Summary
107(2)
References
107(2)
4 Motif Discovery
109(40)
4.1 Motif Discovery Problem(s)
109(1)
4.2 Motif Discovery in Discrete Sequences
110(8)
4.2.1 Projections Algorithm
114(1)
4.2.2 GEMODA Algorithm
115(3)
4.3 Discretization Algorithms
118(6)
4.3.1 MDL Extended Motif Discovery
120(4)
4.4 Exact Motif Discovery
124(10)
4.4.1 MK Algorithm
125(2)
4.4.2 MK+ Algorithm
127(2)
4.4.3 MK++ Algorithm
129(2)
4.4.4 Motif Discovery Using Scale Normalized Distance Function (MN)
131(3)
4.5 Stochastic Motif Discovery
134(2)
4.5.1 Catalano's Algorithm
134(2)
4.6 Constrained Motif Discovery
136(7)
4.6.1 MCFull and MCInc
136(2)
4.6.2 Real-Valued GEMODA
138(1)
4.6.3 Greedy Motif Extension
138(2)
4.6.4 Shift-Density Constrained Motif Discovery
140(3)
4.7 Comparing Motif Discovery Algorithms
143(1)
4.8 Real World Applications
144(2)
4.8.1 Gesture Discovery from Accelerometer Data
144(1)
4.8.2 Differential Drive Motion Pattern Discovery
145(1)
4.8.3 Basic Motions Discovery from Skeletal Tracking Data
145(1)
4.9 Summary
146(3)
References
147(2)
5 Causality Analysis
149(22)
5.1 Causality Discovery
150(1)
5.2 Correlation and Causation
150(1)
5.3 Granger-Causality and Its Extensions
151(2)
5.4 Convergent Cross Mapping
153(9)
5.5 Change Causality
162(3)
5.6 Application to Guided Navigation
165(1)
5.6.1 Robot Guided Navigation
165(1)
5.7 Summary
166(5)
References
166(5)
Part II Autonomously Social Robots
6 Introduction to Social Robotics
171(22)
6.1 Engineering Social Robots
171(3)
6.2 Human Social Response to Robots
174(3)
6.3 Social Robot Architectures
177(13)
6.3.1 C4 Cognitive Architecture
177(4)
6.3.2 Situated Modules
181(4)
6.3.3 HAMMER
185(5)
6.4 Summary
190(3)
References
190(3)
7 Imitation and Social Robotics
193(14)
7.1 What Is Imitation?
193(3)
7.2 Imitation in Animals and Humans
196(4)
7.3 Social Aspects of Imitation in Robotics
200(4)
7.3.1 Imitation for Bootstrapping Social Understanding
201(1)
7.3.2 Back Imitation for Improving Perceived Skill
202(2)
7.4 Summary
204(3)
References
204(3)
8 Theoretical Foundations
207(22)
8.1 Autonomy, Sociality and Embodiment
207(4)
8.2 Theory of Mind
211(7)
8.3 Intention Modeling
218(6)
8.3.1 Traditional Intention Modeling
219(2)
8.3.2 Intention in Psychology
221(1)
8.3.3 Challenges for the Theory of Intention
222(1)
8.3.4 The Proposed Model of Intention
223(1)
8.4 Guiding Principles
224(1)
8.5 Summary
225(4)
References
225(4)
9 The Embodied Interactive Control Architecture
229(16)
9.1 Motivation
229(1)
9.2 The Platform
230(3)
9.3 Key Features of EICA
233(1)
9.4 Action Integration
234(3)
9.4.1 Behavior Level Integration
236(1)
9.4.2 Action Level Integration
236(1)
9.5 Designing for EICA
237(1)
9.6 Learning Using FPGA
238(2)
9.7 Application to Explanation Scenario
240(2)
9.7.1 Fixed Structure Gaze Controller
241(1)
9.8 Application to Collaborative Navigation
242(1)
9.9 Summary
243(2)
References
243(2)
10 Interacting Naturally
245(10)
10.1 Main Insights
245(2)
10.2 EICA Components
247(2)
10.3 Down-Up-Down Behavior Generation (DUD)
249(3)
10.4 Mirror Training (MT)
252(1)
10.5 Summary
253(2)
References
253(2)
11 Interaction Learning Through Imitation
255(20)
11.1 Stage 1: Interaction Babbling
255(4)
11.1.1 Learning Intentions
256(1)
11.1.2 Controller Generation
257(2)
11.2 Stage 2: Interaction Structure Learning
259(7)
11.2.1 Single-Layer Interaction Structure Learner
259(2)
11.2.2 Interaction Rule Induction
261(3)
11.2.3 Deep Interaction Structure Learner
264(2)
11.3 Stage 3: Adaptation During Interaction
266(3)
11.3.1 Single-Layer Interaction Adaptation Algorithm
266(2)
11.3.2 Deep Interaction Adaptation Algorithm
268(1)
11.4 Applications
269(3)
11.4.1 Explanation Scenario
270(1)
11.4.2 Guided Navigation Scenario
271(1)
11.5 Summary
272(3)
References
272(3)
12 Fluid Imitation
275(18)
12.1 Introduction
276(2)
12.2 Example Scenarios
278(1)
12.3 The Fluid Imitation Engine (FIE)
279(1)
12.4 Perspective Taking
280(6)
12.4.1 Transforming Environmental State
280(2)
12.4.2 Calculating Correspondence Mapping
282(4)
12.5 Significance Estimator
286(2)
12.6 Self Initiation Engine
288(1)
12.7 Application to the Navigation Scenario
288(2)
12.8 Summary
290(3)
References
290(3)
13 Learning from Demonstration
293(26)
13.1 Early Approaches
294(1)
13.2 Optimal Demonstration Methods
295(12)
13.2.1 Inverse Optimal Control
295(4)
13.2.2 Inverse Reinforcement Learning
299(3)
13.2.3 Dynamic Movement Primitives
302(5)
13.3 Statistical Methods
307(6)
13.3.1 Hidden Markov Models
307(1)
13.3.2 GMM/GMR
307(6)
13.4 Symbolization Approaches
313(3)
13.5 Summary
316(3)
References
316(3)
14 Conclusion
319(6)
Index 325