Stock market prediction using an improved training algorithm of neural network

A comprehensive evaluation of ensemble learning for stock-market prediction

“…We observed that the DT ensemble classifiers by boosting (DTBotc) and bagging (DTBagc) obtain an accuracy of 99.98% with (10-200) estimators over the GSE, BSE, and NYSE dataset (Figs. 6,7,8,9). The accuracy of the MLP ensemble by bagging Likewise, it was observed that the DT ensemble classifiers DTBotc and DTBagc performed very-well over NYSE at 100% accuracy with (1-20) estimators as compared with the JSE, GSE, and BSE.…”

“…To overcome the challenges in the stock market analysis, several computational models based on soft-computing and machine learning paradigms have been used in the stock-market analysis, prediction, and trading. Techniques like Support Vector Machine (SVM) [2,5], DTs [6], neural networks [7], Naïve Bayes [8,9] and artificial neural networks (ANN) [10,11] were reported to have performed better in stockmarket prediction than conventional arithmetic methods like Logistic regression (LR), in respect of error prediction and accuracy. Nevertheless, ensemble learning (EL) based on a learning-paradigm that combines multiple learning algorithms, forming committees to improve-predictions (stacking and blending) or decrease variance (bagging), and bias (boosting) is believed to perform better than single classifiers and regressors [12,13].…”

A Novel Ensemble Neuro-Fuzzy Model for Financial Time Series Forecasting

“…Examples of successfully applied ANFIS-derived models for stock market prediction are the neuro-fuzzy model with a modification of the Levenberg-Marquardt learning algorithm for Dhaka Stock Exchange day closing price prediction [3] and an ANFIS model based on an indirect approach and tested on Tehran Stock Exchange Indexes [4].…”

Recurrent Neural Networks for Financial Time-Series Modelling

“…Similar to Hidden Markov Models (HMMs), such recurrent architectures by nature are suitable for modelling sequential data with delayed temporal correlations. Many recent works with RNN family have shown good promise in stock price prediction using either technical indicators [10] or social sentiments [11].…”