Classification of trading strategies of agents in a competitive market

Transcription

1 Classification of trading strategies of agents in a competitive market CS Machine Learning Final Project presentation Mark Gruman Manjunath Narayana 12/12/27

2 Application CAT tournament Objective Facilitate transactions between buyers and sellers based on their expected profit margins Maximize our profit from transactional fees (static) and commission (based on transaction price) Why perform classification? Knowing which strategies are utilized can help us adjust our fees to maximize transactions and profit

3 Trading strategies used by agents 4 known strategies Double Auction (Gjerstad-Dickhaut) New bids depend on expectation of profit from bid Keeps track of what fraction of bids at a particular price were accepted Extensive-Form game (Roth-Erev) New bids based on profits from previous bids in the system Keeps track of how much profit resulted from each bid ZI-C: Zero Information, Constrained (Gode-Sunder) New random bids, below value (buyers) and above cost (sellers) ZIP: Zero Information Plus (Dave Cliff) New random bids, below value (buyers) and above cost (sellers) Keeps track of other bids and adjusts margins based on market

4 Data generation Buyers - GD Buyers - GD Buyers - GD Buyers - GD 1,12,3,4,23,45,4,5,6 MARKET 21,2,4,3,45,7,2,4,33 Sellers - GD Sellers - GD Sellers - GD Sellers - GD

5 Data Collection Issues 4 strategies multiclass problem Many traders use same strategy We don t know which trader uses which strategy Bids are masked before they reach the market We don t know which bid belongs to which trader If a bid is accepted, the string of bids terminates, and the same trader begins a new series of bids using a new bid ID Bids are being shouted in no particular order Some traders bid much more often than others Bids are adjusted based on market response A number of data conversion tools had to be developed

6 Illustration of data Single seller example Bids can be of any length Depends on the price buyers are willing to buy at Seller_GD_1-84., 9.4, 12.9 Seller_GD_ Seller_GD_ Seller_GD_1-59., 85.2, 82.7, 81.3, 73.4, can be a large number Overall Market Bid can be any length, any seller belonging to any strategy Actual identities are hidden Seller_GD_1-84., 9.4, 12.9 Seller_RE_ Seller_GD_ , 14.5, 75.6, 9.3 Seller_ZIP_1-1.2, 98.7, 89.6, 77.6, 19.4 Seller_ZIP_4-59., 85.2, 82.7, 81.3, 73.4 Seller_ZIC_3-34.5,56.7,78.9 Seller_GD_ , 132,6

7 Classification Generate data where the identities are not hidden Divide into classes (GD,RE,ZIP,ZIC) Break into training and testing sets Learn on training set Classify the testing set

8 K-means Results Collected data did not fit the k-means model Different number of bid sequences per strategy Variable size of bid sequences, had to limit to 1 Overall Result: GD RE ZIP ZIC GD RE ZIP ZIC Accuracy = 22.4%

9 K-means Plot

10 SVM Results GD RE ZIP ZIC SVM Linear kernel Accuracy = 32.8% SVM Radial kernel Default gamma=1 SVM Polynomial kernel Default gamma=1, coeff=, degree=3 GD 92 1 RE ZIP ZIC Accuracy = 38.4% SVM Sigmoid kernel Default gamma=1, coeff= GD RE ZIP ZIC GD RE ZIP ZIC GD GD 192 RE RE 135 ZIP ZIP 17 ZIC ZIC 195 Accuracy = 53.8% Accuracy = 28.2%

11 HMM M Q Hidden State HMM 1 States, 1 Mixtures GD RE Mixture parameters X ZIP ZIC Accuracy = 62.28% The Model Bid Best result from multiple runs

12 HMM Results HMM 4 States, 2 Mixtures GD RE ZIP ZIC Accuracy = 52.31% HMM 4 States, 1 Mixtures GD RE ZIP ZIC Accuracy = 55.49% HMM 1 States, 1 Mixture GD RE ZIP ZIC Accuracy = 52.75% HMM 4 States, 1 Mixtures GD RE ZIP ZIC Accuracy = 6.98%

13 Feature reduction Since there only some samples with large number of features (bids), what happens when we truncate the bids to smaller sequences? 2 features Reduction in accuracy 1 features - Improvement (slight) in accuracy HMM First 2 features only GD RE ZIP ZIC Accuracy = 49.71% HMM First 1 features only GD RE ZIP ZIC Accuracy = 62.57%

14 Other methods Other classification techniques: CRF SVM with user defined kernel Fourier kernel Results 38-42% accuracy Would like to try the pyramid kernel Additional information may be obtained via CAT framework More traders, longer training data collection intervals Information via subscription to other markets

15 Summary and Conclusion Difficult dataset Best accuracy with HMM Hidden variables (identity of traders, parameters) Time-series Thus, HMM is the best method SVM also fairly successful (relative) Time varying data Varying number of features Most times 1 feature, sometimes as many as 2 Determining strategy employed by each trader may not be necessary May be sufficient to rely on an accurate distribution

16 References Dave Cliff. Minimal-intelligence agents for bargaining behaviours in market-based environments. Technical Report HP-97-91, Hewlett- Packard Research Laboratories, Bristol, England, S. Gjerstad and J. Dickhaut. Price formation in double auctions. Games and Economic Behaviour, 22:1-29, D. K. Gode and S. Sunder. Allocative efficiency of markets with zerointelligence traders: Markets as a partial substitute for individual rationality. The Journal of Political Economy, 11(1): , February A. E. Roth and I. Erev. Learning in extensive-form games: Experimental data and simple dynamic models in the intermediate term. Games and Economic Behavior, 8: , 1995.

17 Thank you! More information on CAT can be found at: Questions/comments?