A hybrid approach to forecasting futures prices with simultaneous consideration of optimality in ensemble feature selection and advanced artificial intelligence

Ghosh, Indranil; Chaudhuri, Tamal Datta; Alfaro, Esteban; Gámez, Matías; García, Noelia

doi:10.1016/j.techfore.2022.121757

Cited by 26 publications

(9 citation statements)

References 53 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The portioning has been made in a forward-looking sequential manner, i.e., the first 80% of data samples from April 1, 2020, are used to train the models, and the leftover 20% of data points are used to test the respective models sequentially. The aforesaid training process has been documented to be effective for complex time series forecasting in literature ( Ghosh et al, 2022 , Ghosh and Datta Chaudhuri, 2022 ). The calibration of both UMAP-LSTM and ISOMAP-GBR frameworks has been accomplished in the training segment and verified in the test segment using the four performance indicators.…”

Section: Resultsmentioning

confidence: 99%

Role of proliferation COVID-19 media chatter in predicting Indian stock market: Integrated framework of nonlinear feature transformation and advanced AI

Ghosh

Alfaro

Gámez

et al. 2023

Expert Systems with Applications

Self Cite

View full text Add to dashboard Cite

Section: Resultsmentioning

confidence: 99%

Role of proliferation COVID-19 media chatter in predicting Indian stock market: Integrated framework of nonlinear feature transformation and advanced AI

Ghosh

Alfaro

Gámez

et al. 2023

Expert Systems with Applications

Self Cite

View full text Add to dashboard Cite

“…The Boruta feature selection method Ghosh et al (2022), which is quite popular in the forecasting literature, was also evaluated and compared process with the FSA method. As seen in Table 9 the regression MSEs of the Boruta method are higher than FSA and accuracy values are lower than FSA.…”

Section: Feature Selection Resultsmentioning

confidence: 99%

“…The authors reported that the proposed model for feature selection significantly improved the forecasting performance. Ghosh et al (2022) proposed an ensemble feature selection algorithm to forecast the stock price using the national stock exchange data in India by assessing the COVID-19 effect. Spot prices, market sentiment, sectoral outlook, crude price volatility, and exchange rate volatility features were used in this study.…”

Section: Related Workmentioning

confidence: 99%

“…This study investigates the predictability of the selected stock and cryptocurrency return, which means profitability, based on ML techniques with and without feature selection as a regression problem, and the binary buy and sell strategies based on the same techniques as a classification problem. The study evaluates several ML techniques, namely Logistic Regression (LR), artificial neural networks (ANN), convolutional neural networks (CNN), and long short-term memory (LSTM), after applying the feature selection with annealing (FSA) Barbu et al (2016) and the Boruta algorithm Ghosh et al (2022) as a benchmark method.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Feature Selection for Forecasting

Pabuçcu¹,

Barbu²

2023

Preprint

View full text Add to dashboard Cite

This work investigates the importance of feature selection for improving the forecasting performance of machine learning algorithms for financial data. Artificial neural networks (ANN), convolutional neural networks (CNN), long-short term memory (LSTM) networks, as well as linear models were applied for forecasting purposes. The Feature Selection with Annealing (FSA) algorithm was used to select the features from about 1000 possible predictors obtained from 26 technical indicators with specific periods and their lags. In addition to this, the Boruta feature selection algorithm was applied as a baseline feature selection method. The dependent variables consisted of daily logarithmic returns and daily trends of ten financial data sets, including cryptocurrency and different stocks. Experiments indicate that the FSA algorithm increased the performance of ML models regardless of the problem type. The FSA hybrid machine learning models showed better performance in 10 out of 10 data sets for regression and 8 out of 10 data sets for classification. None of the hybrid Boruta models outperformed the hybrid FSA models. However, the BOR-CNN model performance was comparable to the best model for 4 out of 10 data sets for regression estimates. BOR-LR and BOR-CNN models showed comparable performance with the best hybrid FSA models in 2 out of 10 datasets for classification. FSA was observed to improve the model performance in both better performance metrics as well as a decreased computation time by providing a lower dimensional input feature space.

show abstract

“…They also find feedback causality between FDI and financial sector development, and financial sector development and economic growth. Ghosh et al (2022) thoroughly explored the detailed dynamics of the futures market in India during normal and new normal time horizons applying appropriate indicators of spot counterparts, sectoral outlook, market sentiment, market fear, and volatility as explanatory variables. An ensemble of machine learning and explainable artificial intelligence-based frameworks suggested the futures prices of stocks belonging to different sectors were indeed predictable and predominantly driven by the spot markets and sectoral outlook.…”

Section: Previous Researchmentioning

confidence: 99%

Financial Sector Development and Real Sector Growth – Association, Spillover and Causality During Pre Covid and Covid Regimes

Chaudhuri¹,

Ghosh

2022

FU Econ Org

View full text Add to dashboard Cite

In this paper, we propose an alternative approach to understanding the relationship between financial sector development and real sector growth in India. We use stock market sectoral indices available on National Stock Exchange (NSE) like Capital Goods Index, FMCG Index, Energy Index, Infra Index, Metal Index, Realty Index, and Auto Index to represent the real sector. To represent the financial sector, we consider Bank Index and Financial Services Index separately. The proposed framework examines the relationships at a granular level to understand the extent of association, spillover, and causality. We also analyze the relationship between the financial sector and the real sector in Pre COVID and COVID periods separately. Our research methodology includes the use of Detrended Cross-Correlation Analysis (DCCA), Wavelet Multiple Correlation (WMC), Wavelet Multiple Cross Correlation (WMCC), Diebold-Yilmaz spillover Framework, and Non-Linear Causality Test. Our granular approach has enabled us to examine the relationships in different periods and we observe that the results change. The intensity of the relationships also is different during the COVID period as compared to Pre COVID period.

show abstract

A hybrid approach to forecasting futures prices with simultaneous consideration of optimality in ensemble feature selection and advanced artificial intelligence

Cited by 26 publications

References 53 publications

Role of proliferation COVID-19 media chatter in predicting Indian stock market: Integrated framework of nonlinear feature transformation and advanced AI

Role of proliferation COVID-19 media chatter in predicting Indian stock market: Integrated framework of nonlinear feature transformation and advanced AI

Feature Selection for Forecasting

Financial Sector Development and Real Sector Growth – Association, Spillover and Causality During Pre Covid and Covid Regimes

Contact Info

Product

Resources

About