Paper Info
Title
A Hybrid Algorithm for Fast Learning Individual Daily Activity Plans for Multiagent Transportation Simulation
Abstract
This paper propose a hybrid learning algorithm based on the competing risk duration model and the cross entropy method for generating complete all-day activity plan in multiagent transportation simulation. We formulate agent's activity scheduling problem as a sequential Markov decision process. By initially generating individual's activity type and duration sequence from empirical data based on the competing risk duration model, the obtained plans can be efficiently improved by reinforcement learning technique towards near-optimal activity plan. We apply the cross entropy method to efficiently learn near-optimal activity plan. The numerical result shows that the proposed method generates consistent daily activity plans for multiagent transportation simulation.
Year
DOI
Venue
2013
10.1109/TAAI.2013.35
TAAI
Keywords
Field
DocType
cross entropy method,scheduling,activity plan generation,agent activity scheduling problem,reinforcement learning,multiagent transportation simulation,competing risk duration model,decision making,learning (artificial intelligence),activity type,transportation,activity scheduling problem,multi-agent systems,fast learning individual daily,daily activity plans,hybrid fast learning algorithm,hybrid algorithm,individual activity type,near-optimal activity plan,simulation,duration sequence,multiagent,complete all-day activity plan,sequential markov decision process,activity plans,entropy,cross entropy,markov processes,reinforcement learning technique,consistent daily activity plan,learning artificial intelligence,multi agent systems
Cross entropy,Markov process,Hybrid algorithm,Scheduling (computing),Computer science,Markov decision process,Cross-entropy method,Multi-agent system,Artificial intelligence,Machine learning,Reinforcement learning
Conference
ISSN
ISBN
Citations 
2376-6816
978-1-4799-2528-5
0
PageRank 
References 
Authors
0.34
1
2
Name
Order
Citations
PageRank
Tai-Yu Ma132.35
Philippe Gerber221.10