C. L. Giles, S. Lawrence, and A. C. Tsoi, "Noisy time series prediction using recurrent neural networks and grammatical inference," Machine Learning, vol. 44, pp. 161--183, 2001.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....algorithms) the goal of our method is to find rules that are readable and understandable by a human expert. Since this is one of the fundamental goals of data mining and knowledge discovery, we have chosen to classify our method as a rule discovery method. A problem similar to ours is tackled in [7], where Giles et al. use recurrent neural networks to predict fluctuations in foreign exchange rates. In addition to the prediction task, their method encompasses the extraction of deterministic finite state automata, which are equivalent to regular expressions. Like most current sequence learning ....

....the prediction task, their method encompasses the extraction of deterministic finite state automata, which are equivalent to regular expressions. Like most current sequence learning methods, the algorithms works with a fixedwidth sliding window. We have tested our method on the same data sets as [7] in Section 3.3. 2 Method To evolve our predictor rules we use genetic programming with the rule encoding scheme referred to in [6] as the Michigan approach, that is, each individual in the population represents a single rule. Since the consequent is a part of the problem definition, each rule ....

[Article contains additional citation context not shown here]

C. Lee Giles, Steve Lawrence, and Ah Chung Tsoi. Noisy time series prediction using a recurrent neural network and grammatical inference. Machine Learning, 44(1/2):161--183, July/August 2001.

....sux trees, nancial indexes, volatility, straddle, options. 1 Introduction It has been shown in the past that quantizing real valued nancial time series into symbolic streams and subsequent use of predictive models on such sequences can be of great bene t in many nancial tasks [1] [2], 3] 4] 5] 6] 7] 8] 9] This is predominantly due to the inherently noisy and non stationary nature of nancial data. Careful quantization can reduce the noise component in the data while preserving the underlying predictable patterns in the stochastic process. However, the question ....

....For example, Papageorgiou [3] quan2 tized daily returns of exchange rates of ve major currencies into 9 intervals (symbols) No clue was provided as to why 9 quantization intervals were used and how the particular cut values de ning the intervals were chosen. Giles, Lawrence and Tsoi [1] [2] considered the same set of exchange rates. The returns were quantized using one dimensional self organizing feature map [10] without a direct control over the cut values. Up to 7 quantization intervals were considered. We have introduced a data driven parametric scheme for quantizing real valued ....

[Article contains additional citation context not shown here]

C.L. Giles, S. Lawrence, and A.C. Tsoi, \Noisy time series prediction using a recurrent neural network and grammatical inference," Machine Learning, accepted, 2000.

....underlying patterns exploitable for prediction, and (2) they are, or can be, highly non stationary. For a reliable estimation of pattern recognition or prediction models using data from such time series, due to property (1) large samples i.e. time series values over a large time window ( 2] [3]) are needed. Due to property (2) however, the size of such samples cannot arbitrarily be enlarged. We therefore investigate a class of models for prediction which aim at reducing the noise component while preserving the underlying predictable patterns in the stochastic process. This class of ....

....statistical structure. We demonstrate the viability of such an approach in the domain of nancial time series, namely on the application of predicting the volatility of indexes in the nancial markets. The idea of quantizing nancial time series has already appeared in several studies [2] [3], 7] 8] 9] 10] Papageorgiou built predictive models to determine the direction of change in high frequency Swiss franc U.S. dollar exchange rate (XR) tick data [7] and studied the correlational structure of coupled time series of daily Quantizing real valued time series into symbolic ....

[Article contains additional citation context not shown here]

Giles CL, Lawrence S, Tsoi AC. Noisy time series prediction using a recurrent neural network and grammatical inference. Machine Learning, 2000, accepted. 35

....underlying patterns exploitable for prediction, and (2) they are, or can be, highly non stationary. For a reliable estimation of pattern recognition or prediction models using data from such time series, due to property (1) large samples i.e. time series values over a large time window ( 2] [3]) are needed. Due to property (2) however, the size of such samples cannot arbitrarily be enlarged. We therefore investigate a class of models for prediction which aim at reducing the noise component while preserving the underlying predictable patterns in the stochastic process. This class of ....

....statistical structure. We demonstrate the viability of such an approach in the domain of financial time series, namely on the application of predicting the volatility of indexes in the financial markets. The idea of quantizing financial time series has already appeared in several studies [2] [3], 7] 8] 9] 10] Papageorgiou built predictive models to determine the direction of change in high frequency Swiss franc U.S. dollar exchange rate (XR) tick data [7] and studied the correlational structure of coupled time series of daily XRs for five major currencies measured against the ....

[Article contains additional citation context not shown here]

C.L. Giles, S. Lawrence, and A.C. Tsoi, "Noisy time series prediction using a recurrent neural network and grammatical inference," Machine Learning, accepted, 2000.

No context found.

C. L. Giles, S. Lawrence, and A. C. Tsoi, "Noisy time series prediction using recurrent neural networks and grammatical inference," Machine Learning, vol. 44, pp. 161--183, 2001.

No context found.

C. L. Giles, S. Lawrence, and A. C. Tsoi, "Noisy time series prediction using recurrent neural networks and grammatical inference," Machine Learning, vol. 44, pp. 161--183, 2001.

No context found.

L. C. Giles, , S. Lawrence, and A. H. Tsoi, Noisy time series prediction using a recurrent neural network and grammatical inference, Machine Learning 44 (2001), no. 1/2, 161--183.

No context found.

Giles, L. C., , Lawrence, S. & Tsoi, A. H. [2001] "Noisy time series prediction using a recurrent neural network and grammatical inference", Machine Learning, 44(1/2), 161--183. Statistics, 28, 100--108.

No context found.

Giles, C. L., Lawrence, S. & Tsoi, A. C. (2001), `Noisy time series prediction using a recurrent neural network and grammatical inference', Machine Learning 44(1/2), 161--183.

No context found.

L. C. Giles, , S. Lawrence, and A. H. Tsoi, Noisy time series prediction using a recurrent neural network and grammatical inference, Machine Learning 44 (2001), no. 1/2, 161--183.

No context found.

L. C. Giles, , S. Lawrence, and A. H. Tsoi, Noisy time series prediction using a recurrent neural net- work and grammatical inference, Machine Learning 44 (2001), no. 1/2, 161--183.

No context found.

Giles CL, Lawrence S, Tsoi AC.(2001) Noisy time series prediction using recurrent neural networks and grammatical inference. Machine Learning 44(1/2): 161-183(23)

No context found.

L. C. Giles, , S. Lawrence, and A. H. Tsoi, Noisy time series prediction using a recurrent neural network and grammatical inference, Machine Learning 44 (2001), no. 1/2, 161--183.

No context found.

Giles, C. L., Lawrence, S., and Tsoi, A. C. (2001). Noisy time series prediction using recurrent neural networks and grammatical inference. Machine Learning, 44(1--2):161--183.

No context found.

L. C. Giles, , S. Lawrence, and A. H. Tsoi, Noisy time series prediction using a recurrent neural network and grammatical inference, Machine Learning 44 (2001), no. 1/2, 161--183.

No context found.

Giles C L, Lawrence S and Tsoi A C. "Noisy Time Series Prediction using a Recurrent Neural Network and Grammatical interface", Machine Learning, Vol.44, Num. 1/2, July/August, (2001), 161-183.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC