Quantitative Neural Network Models - page 2

New comment
 
These notes explore the use of Sydney Lamb’s relational network notion for linguistics to represent the logical structure of complex collection of attractor landscapes (as in Walter Freeman’s account of neuro-dynamics). Given a sufficiently large system, such as a vertebrate nervous system, one might want to think of the attractor net as itself being a dynamical system, one at a higher order than that of the dynamical systems realized at the neuronal level. A mind is a fluid attractor net of fractional dimensionality over a neural net whose behavior displays complex dynamics in a state space of unbounded dimensionality. The attractor-net moves from one discrete state (frame) to another while the underlying neural net moves continuously through its state space.
 
seekers_:
Interesting
 
In the last decade, neural networks have drawn noticeable attention from many computer and operations researchers. While some previous studies have found encouraging results with using this artificial intelligence technique to predict the movements of established financial markets, it is interesting to verify the persistence of this performance in the emerging markets. These rapid growing financial markets are usually characterized by high volatility, relatively smaller capitalization, and less price efficiency, features which may hinder the effectiveness of those forecasting models developed for established markets. In this study, we attempt to model and predict the direction of return on the Taiwan Stock Exchange Index, one of the fastest growing financial exchanges in developing Asian countries. Our approach is based on the notion that trading strategies guided by forecasts of the direction of price movement may be more effective and lead to higher profits. The Probabilistic Neural Network (PNN) is used to forecast the direction of index return after it is trained by historical data. The forecasts are applied to various index trading strategies, of which the performances are compared with those generated by the buy and hold strategy, and the investment strategies guided by the forecasts estimated by the random walk model and the parametric Generalized Methods of Moments (GMM) with Kalman filter. Empirical results show that the PNN-based investment strategies obtain higher returns than other investment strategies examined in this study. The influences of the length of investment horizon and the commission rate are also considered.
 
Predicting currency movements has always been a problematic task as most conventional econometric models are not able to forecast exchange rates with significantly higher accuracy than a naive random walk model. For large multinational firms which conduct substantial currency transfers in the course of business, being able to accurately forecast the movements of exchange rates can result in considerable improvement in the overall profitability of the firm. In this study, we apply the General Regression Neural Network (GRNN) to predict the monthly exchange rates of three currencies, British pound, Canadian dollar, and Japanese yen. Our empirical experiment shows that the performance of GRNN is better than other neural network and econometric techniques included in this study. The results demonstrate the predictive strength of GRNN and its potential for solving financial forecasting problems.
 
This paper surveys research on Emulative Neural Network (ENN) models as economic forecasters. ENNs are statistical methods that seek to mimic neural processing. They serve as trainable analytical tools that "learn" autonomously. ENNs are ideal for finding non-linear relationships and predicting seemingly unrecognized and unstructured behavioral phenomena. As computing power rapidly progresses, these models are increasingly desirable for economists who recognize that people act in dynamic ways with rational expectations. Unlike traditional regressions, ENNs work well with incomplete data and do not require normal distribution assumptions. ENNs can eliminate substantial uncertainty in forecasting, but never enough to completely overcome indeterminacy.
 
Time series forecasting gets much attention due to its impact on many practical applications. Higher-order neural network with recurrent feedback is a powerful technique which used successfully for forecasting. It maintains fast learning and the ability to learn the dynamics of the series over time. For that, in this paper, we propose a novel model, called Ridge Polynomial Neural Network with Error-Output Feedbacks (RPNN-EOF), which combines three powerful properties: higher order terms, output feedback and error feedback. The well-known Mackey–Glass time series is used to evaluate the forecasting capability of RPNN-EOF. Results show that the proposed RPNN-EOF provides better understanding for the Mackey–Glass time series with root mean square error equal to 0.00416. This error is smaller than other models in the literature. Therefore, we can conclude that the RPNN-EOF can be applied successfully for time series forecasting. Furthermore, the error-output feedbacks can be investigated and applied with different neural network models.
 

This paper is submited for publication on IEEE Transcations for Neural Networks and Learning System (ID: TNNLS-2016-P-6504) This paper presents a novel online adaptive neuro-computing framework for a robust time-series forecast. The proposed framework does mimic the human mind's biological two-thinking model. Our mind makes decisions/calculations using a two-connected system. A first system, System 1, the so-called the intuitive system, makes decisions based on our experience. A second system, System 2, the controller system, does control the decisions of System 1 by either modifying or trusting them. Similarly to the human mind's two-systems model, this paper proposes an artificial framework consisting of two cellular neural network (CNN) systems. The first CNN processor does represent the intuitive system and we call it Intuitive-CNN. The Second CNN processor does represent the controller system, which is called Controller-CNN. Both are connected within a general framework that we name OSA-CNN. The proposed framework is extensively tested, validated and benchmarked with the best state-of-the-art related methods, while involving real field time-series data. Multiple scenarios are considered: traffic flow data extracted from the PeMs traffic database and the 111 time-series collected from the so-called NN3 competitions. The novel OSA-CNN concept does remarkably highly outperform the state-of-the-art competing methods regarding both performance and universality. Index Terms—Time-series forecast (TFS), Cellular neural networks (CNN), Echo state network (ESN), Model reference neural network adaptive control (MRNNAC), Recursive particle swarm optimization (RPSO), Online self-adaptive CNN (OSA-CNN)


 

A novel approach using modular neural networks to forecast exchange rates based on harmonic patterns in Forex market is introduced. The proposed approach employs three algorithms to predict price, validate its prediction and update the system. The model is trained by historical data using major currencies in Forex market. The proposed system's predictions were evaluated by comparing its results with a non-modular neural network. Results showed that the infrastructure market data consist of significant accurate relations that a single network cannot detect these relations and separate trained networks in specific tasks are needed. Comparison of modular and non-modular systems showed that modular neural network outperforms the other one.



 

The paper tackles with local models (LM) for periodical time series (TS) prediction. A novel prediction method is introduced, which achieves high prediction accuracy by extracting relevant data from historical TS for LMs training. According to the proposed method, the period of TS is determined by using autocorrelation function and moving average filter. A segment of relevant historical data is determined for each time step of the TS period. The data for LMs training are selected on the basis of the k-nearest neighbours approach with a new hybrid usefulness-related distance. The proposed definition of hybrid distance takes into account usefulness of data for making predictions at a given time step. During the training procedure, only the most informative lags are taken into account. The number of most informative lags is determined in accordance with the Kraskov's mutual information criteria. The proposed approach enables effective applications of various machine learning (ML) techniques for prediction making in expert and intelligent systems. Effectiveness of this approach was experimentally verified for three popular ML methods: neural network, support vector machine, and adaptive neuro-fuzzy inference system. The complexity of LMs was reduced by TS preprocessing and informative lags selection. Experiments on synthetic and real-world datasets, covering various application areas, confirm that the proposed period aware method can give better prediction accuracy than state-of-the-art global models and LMs. Moreover, the data selection reduces the size of training dataset. Hence, the LMs can be trained in a shorter time.


 

Financial Times Series such as stock price and exchange rates are, often, non-linear and non-stationary. Use of decomposition models has been found to improve the accuracy of predictive models. The paper proposes a hybrid approach integrating the advantages of both decomposition model (namely, Maximal Overlap Discrete Wavelet Transform (MODWT)) and machine learning models (ANN and SVR) to predict the National Stock Exchange Fifty Index. In first phase, the data is decomposed into a smaller number of subseries using MODWT. In next phase, each subseries is predicted using machine learning models (i.e., ANN and SVR). The predicted subseries are aggregated to obtain the final forecasts. In final stage, the effectiveness of the proposed approach is evaluated using error measures and statistical test. The proposed methods (MODWT-ANN and MODWT-SVR) are compared with ANN and SVR models and, it was observed that the return on investment obtained based on trading rules using predicted values of MODWT-SVR model was higher than that of Buy-and-hold strategy.


123
New comment