A hybrid model based on adaptive-network-based fuzzy inference system to forecast Taiwan stock market

Wei, Liang-Ying; Chen, Tai‐Liang; Ho, Tien-Hwa

doi:10.1016/j.eswa.2011.04.127

Cited by 39 publications

(31 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…One of the advantages of SC method is that it is not needed to estimate the clusters [32].Yuan et al [33] have predicted the quality of software in a research via using SC method. Wei et al [34] have developed a model for prediction of the stock market of Taiwan by SC method. Malhotra and Sharma [35] have examined and modelled the relationship between nine independent variables such as the number of words in web pages, page size, the number of tables, graphs in the page, etc.…”

Section: Subtractive Clustering Methodsmentioning

confidence: 99%

Qualifying Articles of Persian Wikipedia Encyclopedia Through J48 Algorithm, ANFIS and Subtractive Clustering

Seyedsadr

Afsharkazemi²,

Nikoomaram³

2015

ACIS

View full text Add to dashboard Cite

Abstract:Since Wikipedia encyclopedia is one of the most popular web sites on the internet, providing accurate information is of abundant importance. In this research, the effective variables on quality of Persian articles are identified and a system is, then, designed for judging articles in three quality levels: high quality, cleanup needed, and deletion. First, the variables relating to the articles included in the list of featured articles, good articles, cleanup needed, and deletion articles are collected. Then, two methods are used for the analysis of data: First, a decision tree explains the relationships among the collected variables as rules that are implemented by adaptive neuro fuzzy interference system. Second, the data are implemented by subtractive clustering algorithm and the error of both methods is, finally, measured and compared. The results indicate that the average daily hits, total views, page length, total number of edits, total number of authors, and number of templates used are directly related to quality of Persian articles while the number of recent number of authors is inversely related to quality of articles.

show abstract

Section: Subtractive Clustering Methodsmentioning

confidence: 99%

Qualifying Articles of Persian Wikipedia Encyclopedia Through J48 Algorithm, ANFIS and Subtractive Clustering

Seyedsadr

Afsharkazemi²,

Nikoomaram³

2015

ACIS

View full text Add to dashboard Cite

show abstract

“…The input of the data included the input data and output data in the form of a data array (Chen et al 2010(Chen et al : 1187. The final action at this stage involved defining and partitioning the universe of discourse for the input variables using the subtractive clustering method (Cakmakci 2007;Wei et al 2011). The next step involved is generating the fuzzy inference system (FIS) (Chen et al 2010;Efendigil et al 2009).…”

Section: Anfis Processmentioning

confidence: 99%

“…where 2 4 / r α = , r is the radius defining a W i neighborhood, and ·   denotes the Euclidean distance (Wei et al 2011). The data point with many neighbouring data points is chosen as the first cluster centre.…”

Section: Anfis Architecturementioning

confidence: 99%

“…The FIS parameters from the training datasets were then optimised, using the least square method and the backpropogation gradient descent method for training the forecasting ANFIS models (Wei et al 2011;yetilmezsoy et al 2011). The training of the study's data was performed automatically in the ANFIS system and an array of training errors was obtained (Chen et al 2010(Chen et al : 1187.…”

Section: Anfis Processmentioning

confidence: 99%

See 1 more Smart Citation

An Adaptive Neuro-Fuzzy Inference System for Forecasting Australia’s Domestic Low Cost Carrier Passenger Demand

2015

View full text Add to dashboard Cite

Abstract. This study has proposed and empirically tested two Adaptive Neuro-Fuzzy Inference System (ANFIS) models for the first time for predicting Australia's domestic low cost carriers' demand, as measured by enplaned passengers (PAX Model) and revenue passenger kilometres performed (RPKs Model). In the ANFIS, both the learning capabilities of an artificial neural network (ANN) and the reasoning capabilities of fuzzy logic are combined to provide enhanced prediction capabilities, as compared to using a single methodology. Sugeno fuzzy rules were used in the ANFIS structure and the Gaussian membership function and linear membership functions were also developed. The hybrid learning algorithm and the subtractive clustering partition method were used to generate the optimum ANFIS models. Data was normalized in order to increase the model's training performance. The results found that the mean absolute percentage error (MAPE) for the overall data set of the PAX and RPKs models was 1.52% and 1.17%, respectively. The highest R 2 -value for the PAX model was 0.9949 and 0.9953 for the RPKs model, demonstrating that the models have high predictive capabilities.

show abstract

“…Fusion techniques have been proposed like the consolidation of neural networks with fuzzy systems [1]. Some constructs may well include fuzzy neural networks FNN [9], adaptive network fuzzy information system ANFIS [10] or wavelet fuzzy neural networks WLFNN [11]. Neural networks can be enhanced through certain methods such as the genetic algorithm, particle swarm optimization [12] or bacterial foraging optimization [13] so that in order to boost the accuracy of forecasting the best set of parameters can be gained.…”

Section: Introduction and Literature Reviewmentioning

confidence: 99%

Effects of Technical Market Indicators on Stock Market Index Direction Forecasting

Şenyurt¹,

Subaşı²

2015

Modeling of Artificial Intelligence

View full text Add to dashboard Cite

In this study, we examine and compare the performances of data mining techniques based on the daily indexes of five different markets, namely DAX, FTSE 100, NASDAQ Composite, NIKKEI 225 and NYSE Composite. The A/D oscillator, chaikin oscillator, moving average convergence/divergence, stochastic oscillator, acceleration, momentum, chaikin volatility, fast stochastics, slow stochastics, williams %R, negative volume index, positive volume index, relative strength index, accumulation/distribution line, bollinger band, highest high, lowest low, median price, on balance volume, price rate of change, price-volume trend, typical price, volume rate of change, weighted close and williams accumulation/distribution make up the 28 technical indicator attributes used in this study. The data mining techniques utilized in this study comprise of random forest model, artificial neural network (ANN) and support vector machine (SVM). Initially, the input features are sorted with WEKA's Info Ranker which reveals the best attributes based on relative weight. Afterwards, ten attributes having the highest relative weight values are fed into random forest, ANN and SVM. Then different attribute combinations based on their relative weights are tested in turns for more precision.

show abstract

A hybrid model based on adaptive-network-based fuzzy inference system to forecast Taiwan stock market

Cited by 39 publications

References 19 publications

Qualifying Articles of Persian Wikipedia Encyclopedia Through J48 Algorithm, ANFIS and Subtractive Clustering

Qualifying Articles of Persian Wikipedia Encyclopedia Through J48 Algorithm, ANFIS and Subtractive Clustering

An Adaptive Neuro-Fuzzy Inference System for Forecasting Australia’s Domestic Low Cost Carrier Passenger Demand

Effects of Technical Market Indicators on Stock Market Index Direction Forecasting

Contact Info

Product

Resources

About