The DOI matches its Crossref record; 61 of the 64 references it cites were found in a registry or an index; 6 of the 55 sources its claims cite were checked for support.
A comprehensive evaluation of ensemble learning for stock-market prediction
Open accessAbstract
Abstract Stock-market prediction using machine-learning technique aims at developing effective and efficient models that can provide a better and higher rate of prediction accuracy. Numerous ensemble regressors and classifiers have been applied in stock market predictions, using different combination techniques. However, three precarious issues come in mind when constructing ensemble classifiers and regressors. The first concerns with the choice of base regressor or classifier technique adopted. The second concerns the combination techniques used to assemble multiple regressors or classifiers and the third concerns with the quantum of regressors or classifiers to be ensembled. Subsequently, the number of relevant studies scrutinising these previously mentioned concerns are limited. In this study, we performed an extensive comparative analysis of ensemble techniques such as boosting, bagging, blending and super learners (stacking). Using Decision Trees (DT), Support Vector Machine (SVM) and Neural Network (NN), we constructed twenty-five (25) different ensembled regressors and classifiers. We compared their execution times, accuracy, and error metrics over stock-data from Ghana Stock Exchange (GSE), Johannesburg Stock Exchange (JSE), Bombay Stock Exchange (BSE-SENSEX) and New York Stock Exchange (NYSE), from January 2012 to December 2018. The study outcome shows that stacking and blending ensemble techniques offer higher prediction accuracies (90–100%) and (85.7–100%) respectively, compared with that of bagging (53–97.78%) and boosting (52.7–96.32%). Furthermore, the root means square error (RMSE) recorded by stacking (0.0001–0.001) and blending (0.002–0.01) shows a better fit of ensemble classifiers and regressors based on these two techniques in market analyses compared with bagging (0.01–0.11) and boosting (0.01–0.443). Finally, the results undoubtedly suggest that an innovative study in the domain of stock market direction prediction ought to include ensemble techniques in their sets of algorithms.
Overview
CoChat Summary
TL;DR
This paper compares bagging, boosting, stacking, blending, and max voting ensembles built from decision trees, SVM, and MLP for stock-market classification and regression across four exchanges. It reports that stacking and blending achieve the highest accuracy and lowest error, though at higher computational cost.
Key findings
- Stacking and blending offered higher prediction accuracies: stacking 90–100%, blending 85.7–100%, compared with bagging 53–97.78% and boosting 52.7–96.32%; RMSE was stacking 0.0001–0.001, blending 0.002–0.01, bagging 0.01–0.11, and boosting 0.01–0.443.
- Stacking outperformed all other combination techniques for ensemble classifiers and regressors; blending followed, then DT ensembles by boosting/bagging. DT ensemble classifiers by boosting and bagging obtained 99.98% accuracy with 10–200 estimators over GSE, BSE, and NYSE; DT boosting averaged 100% over NYSE, 95.09% over JSE, 98.52% over GSE, and 99.98% over BSE.
- MLP bagging classifier performed better than MLP boosting: MLP bagging recorded 100% over NYSE, 81.53% over JSE, 97.98% over GSE, and 98.93% over BSE, while MLP boosting recorded 96.32% over NYSE, 62.45% over JSE, 88.99% over GSE, and 96.45% over BSE. SVM ensemble classification was low, especially on JSE (54.98% bagging; 53.02% boosting), but SVM regression had better error metrics.
- Performance depended on dataset origin; stacking/blending were computationally expensive, while a DT ensemble of 50–100 estimators by boosting was suggested as a low-cost classifier baseline.
Methods
- Built 25 ensemble classifiers/regressors using DT, SVM, and MLP base learners with bagging, boosting, stacking, blending, and max voting; implemented with Scikit-learn, mlens, and Python.
- Data: market indices from GSE, JSE, NYSE, and BSE-SENSEX; the abstract says January 2012–December 2018, while Table 2 lists GSE 1100 (Jan 2012–Dec 2017, 9 features), NYSE 1760 (Jan 03 2012–Dec 2018, 15 features), JSE 1749 (Jan 2012–Dec 2018, 7 features), and BSE 984 (Jan 2015–Dec 2018, 12 features). The study predicted 30-day-ahead closing price (regression) and price movement (classification).
- Preprocessing used wavelet transform denoising and min-max normalization; evaluation used 10-fold cross-validation, an 80/20 train/test split, and 12 accuracy/error metrics.
Limitations
- Stated: the study predicts stock market indices, not exact stock prices; authors propose future work on feature selection and SVM parameter optimization (e.g., GA, PCA).
- Stated: classical SVM assumes all features contribute equally and has difficulty choosing suitable parameters (C, σ, ε), affecting practicality; SVM ensemble classification was deficient with input variables and dataset sizes.
- Evident from the text: dataset periods in Table 2 do not fully match the abstract (GSE ends Dec 2017; BSE starts Jan 2015 vs abstract Jan 2012–Dec 2018), and some table entries appear inconsistent/OCR-corrupted (e.g., MLPBotc 150 mean reported as 0.135 in the BSE table), limiting reproducibility.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
5 Valid · 4 Not valid · 17 Not checked · 120 Own results and interpretations
The sources with a verdict support this claim; the rest weren’t checked yet, have no full text we could read, or their verdicts aren’t rated yet.
In the paper
“Despite numerous works revealing the dominance of ensemble classifier over single classifier, most of these studies only ensemble a specific type of classifier or regressor for stock-market prediction, such as NN [18][19][20], DT [21,22] and SVM [12,23].”
§ A comprehensive evaluation of ensemble learning for stock-market prediction > Introduction
What the sources say
- Not matchedcites [18] Macchiarulo A. Predicting and beating the stock market with machine learning and technical analysis. J Intern Bank Commer. 2018;23:1–22.
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
- Not matchedcites [21] Weng B. Application of machine learning techniques for stock market prediction. Auburn: Auburn University; 2017.
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
- Supportedcites [22] Khaidem L, Saha S, Dey SR. Predicting the direction of stock market prices using random forest. Appl Math Financ. 2016;2016:1–20. source
The citation supports the decision-tree example in the citing sentence, not its broader claim about most previous studies. In Related Work (p. 2), the source describes its random forest as a multitude of decision trees whose outputs are combined. (LLM, conf=0.92)
- Not checked — no full text of the sourcecites [12] Ballings M, Van den Poel D, Hespeels N, Gryp R. Evaluating multiple classifiers for stock price direction prediction. Expert Syst Appl. 2015;42:7046–56. https://doi.org/10.1016/j.eswa.2015.05.013. source
- Not checked — no full text of the sourcecites [23] Gonzalez TR, Padilha AC, Couto AD. Ensemble system based on genetic algorithm for stock market forecasting. In: 2015 IEEE congress on evolutionary computation. 2015. pp. 3102–8. source
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Not matched
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“most of the studies compared between ensemble classifiers [12, 18-20, 22, 23, 28] or regressors [21,30] machine learning algorithms, but not both.”
§ A comprehensive evaluation of ensemble learning for stock-market prediction > Related works evaluation
What the sources say
- Not checked — no full text of the sourcecites [12] Ballings M, Van den Poel D, Hespeels N, Gryp R. Evaluating multiple classifiers for stock price direction prediction. Expert Syst Appl. 2015;42:7046–56. https://doi.org/10.1016/j.eswa.2015.05.013. source
- Not matchedcites [18] Macchiarulo A. Predicting and beating the stock market with machine learning and technical analysis. J Intern Bank Commer. 2018;23:1–22.
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
- Not checked — no full text of the sourcecites [19] Mabu S, Obayashi M, Kuremoto T. Ensemble learning of rule-based evolutionary algorithm using multi-layer perceptron for supporting decisions in stock trading problems. Appl Soft Comput. 2015;36:357–67. https://doi.org/10.1016/j.asoc.2015.07.020. source
- Not checked yetcites [20] Maknickiene N, Lapinskaite I, Maknickas A. Application of ensemble of recurrent neural networks for forecasting of stock market sentiments. Equilib Q J Econ Econ Policy. 2018;13:7–27. https://doi.org/10.24136/eq.2018.001. source
Checked before — its earlier verdict is held back until a new check runs.
- Partly supportedcites [22] Khaidem L, Saha S, Dey SR. Predicting the direction of stock market prices using random forest. Appl Math Financ. 2016;2016:1–20. sourceExperimental — accuracy not yet measured
Section 3.5 (p. 8) identifies the paper’s random forest as an ensemble classifier, and Section 7 (p. 16) compares its results with other classification models. It does not compare multiple ensemble classifiers, as the citing sentence suggests; the named comparators include logistic regression, discriminant analysis, and SVM. (LLM, conf=0.83)
- Not checked — no full text of the sourcecites [23] Gonzalez TR, Padilha AC, Couto AD. Ensemble system based on genetic algorithm for stock market forecasting. In: 2015 IEEE congress on evolutionary computation. 2015. pp. 3102–8. source
- Not checked — no full text of the sourcecites [28] Pulido M, Melin P, Castillo O. Particle swarm optimization of ensemble neural networks with fuzzy aggregation for time series prediction of the Mexican Stock Exchange. Inf Sci (Ny). 2014;342:317–29. https://doi.org/10.1007/978-3-319-32229-2_23. source
- Not matchedcites [21] Weng B. Application of machine learning techniques for stock market prediction. Auburn: Auburn University; 2017.
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
- Not checked — no full text of the sourcecites [30] Booth A, Gerding E, Mcgroarty F. Automated trading with performance weighted random forests and seasonality. Expert Syst Appl. 2014;41:3651–61. https://doi.org/10.1016/j.eswa.2013.12.009. source
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
The cited sources that could be checked couldn’t confirm or dispute this claim.
In the paper
“These metrics were selected due to their appropriateness and effectiveness for classification and regression ML tasks in stock market prediction [1,27,62].”
§ A comprehensive evaluation of ensemble learning for stock-market prediction > Model evaluation
What the sources say
- Not checked — no full text of the sourcecites [1] Nti IK, Adekoya AF, Weyori BA. A systematic review of fundamental and technical analysis of stock market predictions. Artif Intell Rev. 2019. https://doi.org/10.1007/s10462-019-09754-z. source
- Not checked — no full text of the sourcecites [27] Jiang M, Liu J, Zhang L, Liu C. An improved Stacking framework for stock index prediction by leveraging tree-based ensemble models and deep learning algorithms. Phys Stat Mech Appl. 2019. https://doi.org/10.1016/j.physa.2019.122272. source
- Could not verifycites [62] Kamel SR, Yaghoubzadeh R, Kheirabadi M. Improving the performance of support-vector machine by selecting the best features by Gray Wolf algorithm to increase the accuracy of diagnosis of breast cancer. J Big Data. 2019. https://doi.org/10.1186/s40537-019-0247-7. sourceExperimental — accuracy not yet measured
Our copy of Kamel et al. (2019) is entirely about breast-cancer diagnosis: the 'Preferences and criteria for comparison' section defines accuracy, recall, specificity, precision, F-measure, RMSE, MSE, MCC and Kappa (Eqs. 7-15) as the evaluation set for a GWO-SVM classifier on the UCI breast-cancer dataset (699 records, two classes), and ROC/AUC are discussed under 'Accuracy criterion'. No passage addresses stock market prediction or financial regression/classification tasks (searched: 'stock market prediction classification regression evaluation metrics', 'twelve evaluation criteria accuracy precision recall F-measure MAE'), so the paper neither supports nor contradicts the claim that these twelve metrics are appropriate for stock-market ML tasks. The citation appears to be borrowed for its metric list rather than for any stock-market finding. (LLM, conf=0.72)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“97.98% over GSE”
§ A comprehensive evaluation of ensemble learning for stock-market prediction > Homogenous ensembled classifiers by BAG and BOT
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“99.93 over BSE”
§ A comprehensive evaluation of ensemble learning for stock-market prediction > Homogenous ensembled classifiers by BAG and BOT
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
94 more claims
Create a free account to see all 146 claims and to verify the 5 references still open.
- It may state two separate claims as one. 28
- Probably not a substantial claim. 19
- Borderline: one of the tests for a substantial claim is unclear. 1
- The quote only partly states this claim. 3
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Most ensemble classifiers and regressors are developed by combining decision trees, support vector machines, and neural networks (Nti et al. 2020).”
From Forecasting Stock Market Prices Using Machine Learning and Deep Learning Models: A Systematic Review, Performance Analysis and Discussion of Implications (2023) · cites this paper as Nti et al. 2020
Experimental — accuracy not yet measured
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ A comprehensive evaluation of ensemble learning for stock-market prediction > Error metrics analysis of heterogeneous ensembled classifier by STK and BLD
“Oddly, stacking ensemble regressor (STK_DSN_R) outperformed all regressors by stacking and blending over all datasets.”
§ A comprehensive evaluation of ensemble learning for stock-market prediction > Additional information
“STK_DSN_R 0.074 0.061 0.993 0.993 0.057 0.014 0.279 0.004 STK_SND_R 0.210 0.157 0.942 0.942 0.128 0.042 5.945 0.001 STK_DNS_R 0.055 0.050 0.996 0.998 0.051 0.011 4.480 0.001”
Citations
Cites
Tikhonov Regularization and Total Least Squares
Golub, G. H., Hansen, P. C., & O'Leary, D. P. (1999) · SIAM Journal on Matrix Analysis and Applications · doi:10.1137/s0895479897326432 · Cited by 1,029
Chong, E., Han, C., & Park, F. C. (2017) · Expert Systems with Applications · doi:10.1016/j.eswa.2017.04.030 · Cited by 834
Deep neural networks, gradient-boosted trees, random forests: Statistical arbitrage on the S&P 500
Krauss, C., Do, X. A., & Huck, N. (2017) · European Journal of Operational Research · doi:10.1016/j.ejor.2016.10.031 · Cited by 750
Evaluating multiple classifiers for stock price direction prediction
Ballings, M., Van den Poel, D., Hespeels, N., & Gryp, R. (2015) · Expert Systems with Applications · doi:10.1016/j.eswa.2015.05.013 · Cited by 611
A systematic review of fundamental and technical analysis of stock market predictions
Nti, I. K., Adekoya, A. F., & Weyori, B. A. (2020) · Artificial Intelligence Review · doi:10.1007/s10462-019-09754-z · Cited by 547
Self-citationEnsemble methods for wind and solar power forecasting—A state-of-the-art review
Ren, Y., Suganthan, P. N., & Srikanth, N. (2015) · Renewable and Sustainable Energy Reviews · doi:10.1016/j.rser.2015.04.081 · Cited by 435
The Evolution of Boosting Algorithms: From Machine Learning to Statistical Modelling
Binder, H., Gefeller, O., Schmid, M., & Mayr, A. (2014) · Methods of Information in Medicine · doi:10.3414/me13-01-0122 · Cited by 412
Big Data: Deep Learning for financial sentiment analysis
Sohangir, S., Wang, D., Pomeranets, A., & Khoshgoftaar, T. M. (2018) · Journal of Big Data · doi:10.1186/s40537-017-0111-6 · Cited by 390
Chen, Y., & Hao, Y. (2017) · Expert Systems with Applications · doi:10.1016/j.eswa.2017.02.044 · Cited by 382
A deep learning ensemble approach for crude oil price forecasting
Zhao, Y., Li, J., & Yu, L. (2017) · Energy Economics · doi:10.1016/j.eneco.2017.05.023 · Cited by 380
Research works printed but not linked: 3
- Pimprikar R, Ramachadran S, Senthilkumar K. Use of machine learning algorithms and twitter sentiment analysis for stock market prediction. Int J Pure Appl Math. 2017;115:521–6.
- Priya P, Muthaiah U, Balamurugan M. Predicting yield of the crop using machine learning algorithm. Int J Eng Sci Res Technol. 2018;7:1–7.
- Agarwal P, Bajpai S, Pathak A, Angira R. Stock market price trend forecasting using. Int J Res Appl Sci Eng Technol. 2017;5:1673–6.
Cited by
A Survey of Ensemble Learning: Concepts, Algorithms, Applications, and Prospects
Mienye, I. D., & Sun, Y. (2022) · IEEE Access · doi:10.1109/access.2022.3207287 · Cited by 1,298
A comprehensive review on ensemble deep learning: Opportunities and challenges
Mohammed, A., & Kora, R. (2023) · Journal of King Saud University - Computer and Information Sciences · doi:10.1016/j.jksuci.2023.01.014 · Cited by 1,127
Rouf, N., Malik, M. B., Arif, T., Sharma, S., Singh, S., Aich, S., & Kim, H.-C. (2021) · Electronics · doi:10.3390/electronics10212717 · Cited by 273
Performance of Machine Learning Algorithms with Different K Values in K-fold CrossValidation
Department of Computer Science and Informatics, U. O. E. A. N. R., Nti, I. K., Nyarko-Boateng, O., & Aning, J. (2021) · International Journal of Information Technology and Computer Science · doi:10.5815/ijitcs.2021.06.05 · Cited by 270
Self-citationSonkavde, G., Dharrao, D. S., Bongale, A. M., Deokate, S. T., Doreswamy, D., & Bhat, S. K. (2023) · International Journal of Financial Studies · doi:10.3390/ijfs11030094 · Cited by 262
“Most ensemble classifiers and regressors are developed by combining decision trees, support vector machines, and neural networks (Nti et al. 2020).”
Background
Metadata
- Authors
- Nti, I. K., Adekoya, A. F., & Weyori, B. A.
- Affiliations
- University of Energy and Natural Resources (GH)
- Corresponding author
- Nti, Isaac Kofi
- Venue
- Journal of Big Data
- Published
- 2020-03-11
- Volume
- 7
- Issue
- 1
- ISSN
- 2196-1115 (ISSN-L)
- OpenAlex
- W3013734981
- Cited by
- 329
- References
- 62
- Citations by year
- 2026: 28 · 2025: 69 · 2024: 69 · 2023: 55 · 2022: 52 · 2021: 48 · 2020: 8
- FWCI
- 25.91
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- cc-by
- creativecommons.org/licenses/by/4.0 (tdm)
- creativecommons.org/licenses/by/4.0 (vor)
- Locations
- Journal Of Big Data (publishedVersion, cc-by)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Topics
- Social Sciences › Decision Sciences › Management Science and Operations Research › Stock Market Forecasting Methods
- Social Sciences › Decision Sciences › Management Science and Operations Research › Forecasting Techniques and Applications
- Social Sciences › Economics, Econometrics and Finance › Finance › Financial Markets and Investment Strategies
- Keywords
- stock market prediction, ensemble learning, stacking, blending, bagging, boosting, decision tree, support vector machine, artificial neural networks, forecast accuracy comparison
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol and qwen3.8-omni-flash).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: A comprehensive evaluation of ensemble learning for stock-market prediction. https://app.cochat.ai/papers/10.1186/s40537-020-00299-5
- The paper
- Nti, I. K., Adekoya, A. F., & Weyori, B. A. (2020). A comprehensive evaluation of ensemble learning for stock-market prediction. Journal of Big Data. https://doi.org/10.1186/s40537-020-00299-5
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.