Pricing and hedging derivative securities with neural networks: Bayesian regularization, early stopping, and bagging

IEEE Transactions on Neural Networks

Volumn 12, Issue 4, 2001, Pages 726-734

  Gençay, Ramazan ^a   Qi, Min ^b  

^a UNIVERSITY OF WINDSOR   (Canada)

^b KENT STATE UNIVERSITY   (United States)

Author keywords

Bagging; Bayesian regularization; Early stopping; Hedging error; Neural networks (NNs); Option price

Indexed keywords

COMPUTER SIMULATION; CONFORMAL MAPPING; FEEDFORWARD NEURAL NETWORKS; MAXIMUM LIKELIHOOD ESTIMATION; OPTIMIZATION; PROBABILITY DENSITY FUNCTION; RANDOM PROCESSES; STATISTICAL METHODS;

BAGGING; BAYESIAN REGULARIZATION; HEDGING ERROR; OPTION PRICE;

MULTILAYER NEURAL NETWORKS;

EID: 0035391107     PISSN: 10459227     EISSN: None     Source Type: Journal    
DOI: 10.1109/72.935086     Document Type: Article

References (26)

1. A method for learning from hints
2. Learning from hints
3. Financial market applications of learning from hints
4. Heteroskedasticity and autocorrelation consistent covariance matrix estimation
5. Using a financial training criterion rather than a prediction criterion
6. Heuristics of instability in model selection
7. Bagging predictors
8. American options with stochastic dividends and volatility: A nonparameteric investigation
9. Model complexity control for regression using VC generalization bounds
10. Comparing predictive accuracy
11. Gauss-Newton approximation to bayesian learning
12. Pricing and hedging derivatives securities with neural networks and a homogeneity hint
13. Learning translation invariant recognition in a massively parallel network
14. Multilayer feedforward networks are universal approximators
15. Visual pattern recognition by moment invariants
16. A nonparametric approach to pricing and hedging derivative securities via learning networks
17. Bayesian interpolation
18. A simple, positive semidefinite, heteroskedasticity and autocorrelation consistent covariance matrix
19. Improving the learning speed of two-layer neural networks by choosing initial values of the adaptive weights
20. How good are support vector machines
21. Connectionist nonparametric regression: Multilayer feedforward networks can learn arbitrary mappings