463,456 Hits in 5.0 sec

Learning sequences of actions in collectives of autonomous agents

Kagan Tumer, Adrian K. Agogino, David H. Wolpert
2002 Proceedings of the first international joint conference on Autonomous agents and multiagent systems part 1 - AAMAS '02  
BACKGROUND: COLLECTIVE INTEL- LIGENCE In this section, we summarize the portion of COIN theory necessary and sufficient to describe the learning of sequences of actions in a distributed system [24]  ...  We extended previous results on collective intelligence to agents attempting to maximize sequences of actions, and used Q- learning with rewards set by COIN theory.  ... 
doi:10.1145/544741.544832 dblp:conf/atal/TumerAW02 fatcat:yjd6bwj55zfkdokrabu2fbyznq

Learning sequences of actions in collectives of autonomous agents

Kagan Tumer, Adrian K. Agogino, David H. Wolpert
2002 Proceedings of the first international joint conference on Autonomous agents and multiagent systems part 1 - AAMAS '02  
BACKGROUND: COLLECTIVE INTEL- LIGENCE B 8 C D G F h H P I j I t U g e X D G H P c ( C { u ) Y R U I t f ¢ q q t S § W G H P g U w D G F ¢ U w Q V c ¤ W t D G H P c ¤ C n c ¤ a É Ề b B 0 Ë Ú D G F h U y  ...  C h u S ¤ u ¤ U g C © D G I f h I t H P C ¢ u t D G F ¢ c ( I t U x W G U X Y S 3 W s ¢ I b c ¤ f ¢ D G Q V U X W t a d c ¤ W G q i V c ¤ D G F 8 C h S § D t Permission to make digital or hard copies of  ... 
doi:10.1145/544829.544832 fatcat:6rq2zfaibfcfpns3katisq53ki

Unsupervised Discovery of Student Strategies

Benjamin Shih, Kenneth R. Koedinger, Richard Scheines
2010 Educational Data Mining  
Unsupervised learning algorithms can discover models of student behavior without any initial work by domain experts, but they also tend to produce complicated, uninterpretable models that may not predict  ...  student learning.  ...  Stepwise-HMM-Cluster serves two goals at once: it tries to build a collection of HMMs to fit the observed sequences of actions, but also requires that the collection predict student learning gain.  ... 
dblp:conf/edm/ShihKS10 fatcat:md7em4bsmzh23i2vf5yvcuisnq

Personalized next-best action recommendation with multi-party interaction learning for automated decision-making [article]

Longbing Cao, Chengzhang Zhu
2021 arXiv   pre-print
CRN represents multiple coupled dynamic sequences of a customer's historical and current states, responses to decision-makers' actions, decision rewards to actions, and learns long-term multi-sequence  ...  Our study demonstrates the potential of personalized deep learning of multi-sequence interactions and automated dynamic intervention for personalized decision-making in complex systems.  ...  Learning personalized next-best actions has to further model the multi-party interactions for each customer and his decision-maker and learn heterogeneous dynamic multi-sequence couplings.  ... 
arXiv:2108.08846v1 fatcat:bwtxgbbzonda7jfusnparqufey

An Extended Learner Modeling Method to Assess Students' Learning Behaviors

Yi Dong, Gautam Biswas
2017 Educational Data Mining  
sequences in such a way that we can learn more accurate models of students' learning behaviors.  ...  Building accurate models from limited amount of student data is difficult; to address this we develop a methodology that uses Monte Carlo Tree Search methods to boost the initial set of student action  ...  sequences that provides more coverage of the students' learning behaviors.  ... 
dblp:conf/edm/DongB17 fatcat:j4kidnfvp5fszecf6qcovtjjty

Does One More Time Verbalization's Sequence Increases the Progression of Motor Learning

Rim Mekni, Slim Oueslati, Saber Abdellaoui, Anissa Bouassida, Makrem Zghibi
2022 Creative Education  
The addition of a second sequence of verbalization during the same session is more effective for the learning than the practice under condition of a single sequence application.  ...  On the other hand, we consider the differences in the action projects between the language interactions of girls, boys and mixed groups. The observation is used as data collection instruments.  ...  Conflicts of Interest The authors declare no conflicts of interest regarding the publication of this paper.  ... 
doi:10.4236/ce.2022.131020 fatcat:l27ggmm5kfhv5fl2wxqa7iu4fe

Combining Self-Supervised Learning and Imitation for Vision-Based Rope Manipulation [article]

Ashvin Nair, Dian Chen, Pulkit Agrawal, Phillip Isola, Pieter Abbeel, Jitendra Malik, Sergey Levine
2017 arXiv   pre-print
We present a learning-based system where a robot takes as input a sequence of images of a human manipulating a rope from an initial to goal configuration, and outputs a sequence of actions that can reproduce  ...  To perform this task, the robot learns a pixel-level inverse dynamics model of rope manipulation directly from images in a self-supervised manner, using about 60K interactions with the rope collected autonomously  ...  One of the key challenges in self-supervised robot learning is collecting enough data for learning skilled behaviors, since the state and action space for practical manipulation tasks is extremely large  ... 
arXiv:1703.02018v1 fatcat:nzsbty5drrdvtksxhlprmr22m4

LaND: Learning to Navigate from Disengagements [article]

Gregory Kahn, Pieter Abbeel, Sergey Levine
2020 arXiv   pre-print
LaND learns a neural network model that predicts which actions lead to disengagements given the current sensory observation, and then at test time plans and executes actions that avoid disengagements.  ...  Consistently testing autonomous mobile robots in real world scenarios is a necessary aspect of developing autonomous navigation systems.  ...  Predictive Model The learned predictive model takes as input the current observation and a sequence of future actions, and outputs a sequence of future disengagement probabilities.  ... 
arXiv:2010.04689v1 fatcat:4fglx5mhqvgk3phk4kkl3xvkyq

Using social media and learning analytics to understand how children engage in scientific inquiry

June Ahn, Michael Gubbels, Jason Yip, Elizabeth Bonsignore, Tamara Clegg
2013 Proceedings of the 12th International Conference on Interaction Design and Children - IDC '13  
Second, we conducted a case study of SINQ with six children, ages 8-11, and collected log data of their interactions in the app.  ...  We applied learning analytics on this log data using a visual analytic tool called LifeFlow.  ...  their sequence of actions in SINQ.  ... 
doi:10.1145/2485760.2485805 dblp:conf/acmidc/AhnGYBC13 fatcat:sgrvbirm5naufgu32qe6nudk34

Multi-Level Sequence GAN for Group Activity Recognition [article]

Harshala Gammulle, Simon Denman, Sridha Sridharan, Clinton Fookes
2018 arXiv   pre-print
for effective learning of group activities.  ...  Our proposed architecture outperforms current state-of-the-art results for sports and pedestrian based classification tasks on Volleyball and Collective Activity datasets, showing it's flexible nature  ...  Fig. 1 . 1 The proposed Multi-Level Sequence GAN (MLS-GAN ) architecture: (a) G is trained with sequences of person-level and scene-level features to learn an intermediate action representation, an 'action  ... 
arXiv:1812.07124v1 fatcat:fy3tvo4z4rh73aejvhaxfbiswe

Activity Recognition: Linking Low-level Sensors to High-level Intelligence

Qiang Yang
2009 International Joint Conference on Artificial Intelligence  
Activity recognition "sees" what is in the window to predict the locations, trajectories, actions, goals and plans of humans and objects.  ...  In this article, I will give an overview of some of the current activity recognition research works and explore a life-cycle of learning and inference that allows the lowestlevel radio-frequency signals  ...  The goal graph is learned from the training data, consisting of sequences of sensor readings and activity labels, to allow the inference of goals in a collective-classification manner.  ... 
dblp:conf/ijcai/Yang09 fatcat:fcbmme7wtbas5pha7g3ijo6orm

Building leadership capacity in school leadership groups: an action research project

Marit Aas, Kirsten Foshaug Vennebo
2021 Educational Action Research  
Combined with the theory of expansive learning, the theories of critical participatory action research and practice architectures frame the study.  ...  The study identified two essential actions for building leadership capacity in school leadership groups: performing an empirical and historical analysis of the problem space worked on and conducting collective  ...  Theory of expansive learning Questioning is a necessary starting point in Engeström's (2001) sequences of action in an expansive learning circle.  ... 
doi:10.1080/09650792.2021.1934710 fatcat:6auwrolegzdcbjmhx5y4gslbdi

Monte Carlo Tree Search for Scheduling Activity Recognition

Mohamed R. Amer, Sinisa Todorovic, Alan Fern, Song-Chun Zhu
2013 2013 IEEE International Conference on Computer Vision  
For querying an activity in the video, MCTS optimally schedules a sequence of detectors and trackers to be run, and where they should be applied in the space-time volume.  ...  Our videos show a number of co-occurring individual and group activities.  ...  This action sequence is appended to by selecting random actions until reaching a depth of B, resulting in a sequence of B actions.  ... 
doi:10.1109/iccv.2013.171 dblp:conf/iccv/AmerTFZ13 fatcat:nhajpvubrbgbphnu7547vwzkxm

A Large-scale Varying-view RGB-D Action Dataset for Arbitrary-view Human Action Recognition [article]

Yanli Ji, Feixiang Xu, Yang Yang, Fumin Shen, Heng Tao Shen, Wei-Shi Zheng
2019 arXiv   pre-print
To provide data for arbitrary-view action recognition, we newly collect a large-scale RGB-D action dataset for arbitrary-view action analysis, including RGB videos, depth and skeleton sequences.  ...  In total, 118 persons are invited to act 40 action categories, and 25,600 video samples are collected. Our dataset involves more participants, more viewpoints and a large number of samples.  ...  Building bridges between a large collection dataset and test datasets, multi-view action recognition was realized by matching sequences of various viewpoints to data samples of the large collection dataset  ... 
arXiv:1904.10681v1 fatcat:zkl72lgskrdkrolppket7r6pqi

Applying Learning Analytics to Detect Sequences of Actions and Common Errors in a Geometry Game

Manuel J. Gomez, José A. Ruipérez-Valiente, Pedro A. Martínez, Yoon Jeon Kim
2021 Sensors  
The final objective is to facilitate that teachers can understand the sequence of actions and common errors of students using Shadowspect so they can better understand the process, make proper assessment  ...  To address this gap, we seek to provide sequence and process mining metrics to teachers that are easily interpretable and actionable.  ...  Conflicts of Interest: The authors declare no conflict of interest.  ... 
doi:10.3390/s21041025 pmid:33546167 fatcat:3zfv66ai3zcwtbwifzxbhmt65m
« Previous Showing results 1 — 15 out of 463,456 results