The nature of statistical learning theory
The nature of statistical learning theory
A fast quantum mechanical algorithm for database search
STOC '96 Proceedings of the twenty-eighth annual ACM symposium on Theory of computing
New ideas in optimization
An introduction to support Vector Machines: and other kernel-based learning methods
An introduction to support Vector Machines: and other kernel-based learning methods
Advanced Engineering Mathematics: Maple Computer Guide
Advanced Engineering Mathematics: Maple Computer Guide
Quantum computation and quantum information
Quantum computation and quantum information
Time Series Analysis: Forecasting and Control
Time Series Analysis: Forecasting and Control
Evolving the Topology of Hidden Markov Models Using Evolutionary Algorithms
PPSN VII Proceedings of the 7th International Conference on Parallel Problem Solving from Nature
StockMarket Forecasting Using Hidden Markov Model: A New Approach
ISDA '05 Proceedings of the 5th International Conference on Intelligent Systems Design and Applications
Elements of Forecasting
Stock time series forecasting using support vector machines employing analyst recommendations
ISNN'06 Proceedings of the Third international conference on Advances in Neural Networks - Volume Part III
Evolving the structure of hidden Markov models
IEEE Transactions on Evolutionary Computation
Hi-index | 0.01 |
Two crucial problems of overshoot and volatility clustering are encountered in time series prediction frequently, which may give rise to big residual errors and then deteriorate the predictive accuracy. Thus this study introduces a novel scheme to overcome the preceding problems at the same time. First, adaptive support vector regression (ASVR) with fewer data is applied to tackling the overshoot results and its achievement outperforms the traditional prediction methods like ARMAX or grey model. Next, the effect of volatility clustering can be resolved by incorporating nonlinear generalized autoregressive conditional heteroscedasticity (NGARCH) into ASVR to form a hybrid model. Finally, in order to avoid the over-fit or under-fit modeling resulted from quadratic optimization or back-propagation neural network, instead a quantum-based algorithm called quantum minimization (QM) is proposed herein to tune the hybrid model of ASVR and NGARCH to overcome the problems of overshoot and volatility clustering simultaneously due to quantum parallelism finding the optimum in depth. In summary, the proposed method obtains the satisfactory results because of best-fitting the dynamics of time series and thus significantly improving the predictive accuracy.