A Multi-Agent Prediction Market based on Dora Matache*, Raj - PowerPoint PPT Presentation

A Multi-Agent Prediction Market based on Boolean Network Evolution Janyl Jumadinova, A Multi-Agent Prediction Market based on Dora Matache*, Raj Dasgupta Boolean Network Evolution Outline Introduction Research Problem Janyl Jumadinova, Dora Matache*, Raj Dasgupta BN-based Prediction Market Experimental Results C-MANTIC Research Group Conclusion Department of Computer Science *Department of Mathematics University of Nebraska at Omaha, USA IAT 2011 1 / 37

Outline A Multi-Agent Prediction Market based on Boolean Problem: How does a prediction market evolve with Network Evolution Janyl Jumadinova, respect to different market parameters? Dora Matache*, Raj Dasgupta Solution: A multi-agent prediction market that uses Outline Boolean network-based rules to capture the evolution of Introduction Research Problem beliefs of the traders as well as to aggregate the market BN-based price Prediction Market Experimental Results Experimental validation: Conclusion Comparison with existing aggregation technique Evaluation of our prediction market with respect to different market parameters 1 / 37

Prediction Market A Multi-Agent Prediction Market based on Boolean Network Evolution Janyl Jumadinova, Dora Matache*, A Prediction market is Raj Dasgupta a market-based mechanism used to Outline Introduction - combine the opinions on a future event from different Research Problem people and BN-based - forecast the possible outcome of the event based on the Prediction Market Experimental aggregated opinion Results Conclusion 2 / 37

Prediction Market A Multi-Agent Prediction Market based on Boolean Network Evolution Janyl Jumadinova, Dora Matache*, Raj Dasgupta Outline Introduction Research Problem BN-based Prediction Market Experimental Results Conclusion 3 / 37

Prediction Market A Multi-Agent Prediction Market based on Boolean Network Evolution Janyl Jumadinova, Dora Matache*, Raj Dasgupta Outline Introduction Research Problem BN-based Prediction Market Experimental Results Conclusion 4 / 37

Prediction Market A Multi-Agent Prediction Market based on Boolean Network Evolution Janyl Jumadinova, Dora Matache*, Raj Dasgupta Outline Introduction Research Problem BN-based Prediction Market Experimental Results Conclusion 5 / 37

Prediction Market A Multi-Agent Prediction Market based on Boolean Network Evolution Janyl Jumadinova, Dora Matache*, Raj Dasgupta Outline Introduction Research Problem BN-based Prediction Market Experimental Results Conclusion 6 / 37

Prediction Market A Multi-Agent Prediction Market based on Boolean Network Evolution Janyl Jumadinova, Dora Matache*, Raj Dasgupta Outline Introduction Research Problem BN-based Prediction Market Experimental Results Conclusion 7 / 37

Prediction Market A Multi-Agent Prediction Market based on Boolean Network Evolution Janyl Jumadinova, Dora Matache*, Raj Dasgupta Outline Introduction Research Problem BN-based Prediction Market Experimental Results Conclusion 8 / 37

Prediction Market A Multi-Agent Prediction Market based on Boolean Network Evolution Janyl Jumadinova, Dora Matache*, Raj Dasgupta Outline Introduction Research Problem BN-based Prediction Market Experimental Results Conclusion 9 / 37

Prediction Market A Multi-Agent Prediction Market based on Boolean Network Evolution Janyl Jumadinova, Dora Matache*, Raj Dasgupta Outline Introduction Research Problem BN-based Prediction Market Experimental Results Conclusion 10 / 37

Prediction Market A Multi-Agent Prediction Market based on Boolean Network Evolution Janyl Jumadinova, Dora Matache*, Raj Dasgupta Outline Introduction Research Problem BN-based Prediction Market Experimental Results Conclusion 11 / 37

Prediction Market A Multi-Agent Prediction Market based on Boolean Network Evolution Janyl Jumadinova, Dora Matache*, Raj Dasgupta Outline Introduction Research Problem BN-based Prediction Market Experimental Results Conclusion 12 / 37

Prediction Market A Multi-Agent Prediction Market based on Boolean Network Evolution Janyl Jumadinova, Dora Matache*, Raj Dasgupta Outline Introduction Research Problem BN-based Prediction Market Experimental Results Conclusion 13 / 37

Prediction Market Main Features A Multi-Agent Prediction Market based on Boolean A prediction market is run for a real-life unknown event Network Evolution Each event has a finite duration Janyl Jumadinova, Dora Matache*, Raj Dasgupta Each event’s outcome has a security associated with it Traders buy and sell the securities based on their beliefs Outline about the outcome of the event Introduction Research Problem Traders’ beliefs are expressed as probabilities BN-based Prediction Market Market maker aggregates the probabilities from all the Experimental traders into a single probability, market price Results Market price of a security represents the probability of Conclusion the outcome of an event associated with that security happening Traders get paid according to their reported beliefs 14 / 37

Boolean Networks (BNs) A Multi-Agent Prediction Market based on Boolean Network Evolution Janyl Jumadinova, Dora Matache*, Raj Dasgupta The state of the node is either ON ( 1 ) or OFF ( 0 ) Outline Introduction The state of a node is updated according to a Boolean Research Problem rule BN-based Prediction Market The Boolean rule determines state transitions of the Experimental nodes Results Can use BNs to explore the dynamics of the network or Conclusion just some relevant nodes BNs have been used to model various real networks: genetic regulatory networks, strongly disordered systems common in physics and biology 15 / 37

Boolean Networks (BNs) A Multi-Agent Prediction Market based on Boolean Network Evolution Janyl Jumadinova, Dora Matache*, Raj Dasgupta BNs are useful for prediction markets since they reduce the complexity of analyzing a large network Outline Correspondence between Boolean values output by the Introduction Research Problem BN’s rules and the binary outcome of events in a BN-based prediction market Prediction Market Can retain the essential aspects of a prediction market, Experimental Results but easy to understand and manipulate Conclusion 16 / 37

Research Problem Addressed A Multi-Agent Prediction Market based on Boolean Network Evolution Janyl Jumadinova, Develop a Boolean network-based prediction market Dora Matache*, using simple Boolean rules for Raj Dasgupta - updating the beliefs for each of the market’s Outline participants, Introduction - aggregating the participants’ belief information into a Research Problem single market price BN-based Prediction Market The Main Research Question: Experimental Results Under what conditions does a prediction market Conclusion perform the best? 17 / 37

Our Solution Boolean Network-based Prediction Market A Multi-Agent Prediction Market based on Boolean Network Evolution Janyl Jumadinova, Dora Matache*, Raj Dasgupta A multi-agent prediction market based on Boolean Outline Network evolution Introduction Boolean rule for traders’ belief updates Research Problem Boolean rule for calculating the market price BN-based Prediction Market Mathematical model of the states of the traders and the Experimental market maker Results Conclusion 18 / 37

Boolean Network-based Prediction Market A Multi-Agent Prediction Market based on Boolean Network Evolution Janyl Jumadinova, Dora Matache*, Raj Dasgupta Outline Introduction Research Problem BN-based Prediction Market - Our BN-based prediction market consists of: Experimental Results trading agents, Conclusion market maker agent, and information sources external to the market - Each trading agent has a state representing its belief - State can take two values: 1 (believe event will happen) or 0 (believe event will not happen) 19 / 37

Boolean Network-based Prediction Market A Multi-Agent Prediction Market based on Boolean Network Evolution Janyl Jumadinova, Dora Matache*, Raj Dasgupta Outline Introduction Research Problem BN-based Prediction Market Experimental Results Conclusion The diagram shows actions of each entity in one time step 20 / 37

Trading Agents’ Boolean Belief Update A Multi-Agent Prediction Market p r ( t ) - the aggregated market price at trading period t based on Boolean Network Evolution r n ( t ) - the state of the n -th trading agent at trading period t Janyl Jumadinova, Dora Matache*, w n i - the trust that the n -th trading agent holds for the Raj Dasgupta accuracy of the posted market price, its own past belief, Outline and the new information signal it obtains, respectively, Introduction following Golub [2010] Research Problem These trusts are represented as weights w n i ∈ [0 , 1] , BN-based Prediction Market such that w n 1 + w n 2 + w n 3 = 1 Experimental Results B n ( t ) - the information signal received by the n -th trading Conclusion agent at trading period t Information signal is the value of Bernoulli random variable with probability q n of obtaining 1 (positive information) 21 / 37