The DOI matches its Crossref record; 76 of the 83 references it cites were found in a registry or an index; 44 of the 76 sources its claims cite were checked for support.
Forecasting Stock Market Prices Using Machine Learning and Deep Learning Models: A Systematic Review, Performance Analysis and Discussion of Implications
Open accessAbstract
A systematic review, performance analysis and discussion of implications of forecasting stock market prices using machine learning and deep learning models. Implements an ensemble model combining Random Forest, XGBoost, and LSTM for stock price prediction.
Overview
CoChat Summary
TL;DR
This paper reviews machine learning, deep learning, time series, and ensemble methods for stock price forecasting and implements an ensemble “Random Forest + XG-Boost + LSTM” on two Indian chemical-sector stocks. The ensemble achieved the best reported RMSE and R² among the compared models for both stocks.
Key findings
- The proposed ensemble “Random Forest + XG-Boost + LSTM” demonstrated the best performance compared with other distinct algorithms in the paper’s experimental analysis.
- For TAINIWALCHM, the ensemble achieved RMSE 2.0247 and R² 0.9921; for AGROPHOS, it achieved RMSE 1.2658 and R² 0.9897 (Table 2).
- The review concludes that machine learning or deep learning models alone are not sufficient, that ensemble techniques can provide superior performance, and that hyperparameter tuning is crucial for improving stock forecasts.
- The paper states there is still no universal solution to accurately predict stock price or market trend, and AI-based models can fail if not trained efficiently with fresh data.
Methods
- Design: a systematic/comprehensive review of supervised and unsupervised machine learning, ensemble, time series, and deep learning algorithms for stock price prediction and classification, plus an experimental comparative analysis.
- Data: stock data obtained from the Yahoo Finance API for Tainwala Chemicals and Plastics (TAINIWALCHM) from 2014 to 2023 and Agro Phos (AGROPHOS) from 2018 to 2023.
- Main analysis: implemented algorithms in Python and evaluated performance using RMSE and R²; the ensemble used a stacked model with Keras, TensorFlow, and sklearn, 80% training and 20% testing, MSE loss, Adam optimizer, and grid-search hyperparameter tuning.
Limitations
- The paper states that no universal solution exists for accurate stock price or trend prediction, that AI-based models can fail if not trained efficiently with fresh data, and that decisions should not rely only on AI-based price forecasting methods.
- The text makes evident that the experimental comparison is limited to two stocks, TAINIWALCHM and AGROPHOS, both described as falling under the chemical industry market sector.
- The text makes evident that the review’s comparative performance table draws on heterogeneous prior studies with different datasets, metrics, and evaluation settings, so the reported comparisons are not from a single controlled benchmark.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
33 Valid · 11 Not valid · 17 Not checked · 26 Own results and interpretations
At least one cited source doesn’t support this claim or points to a different paper, and none confirmed it — see the verdicts below.
In the paper
“The ensemble algorithm of XG-Boost + LSTM shows the following evaluation parameter: MSE = 3.465 (Vuong et al. 2022).”
§ Forecasting Stock Market Prices Using Machine Learning and Deep Learning Models: A Systematic Review, Performance Analysis and Discussion of Implications > 13 XG-Boost + LSTM
What the sources say
- Disputedcites Vuong et al. 2022 sourceExperimental — accuracy not yet measured
Table 6 in §3.3 Results reports an MSE of 3.465 × 10⁻⁷ for XGBoost + LSTM on the Forex dataset, not 3.465. The citing sentence omits the exponent, changing the reported value by a factor of ten million. (LLM, conf=0.95)
Checked by a reasoning model — may be wrong.
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.
- Disputed
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
“Stock forecasting using LSTM, a unique recurrent neural network (RNN), overcomes long-term dependency (Qiu et al. 2020;Banik et al. 2022).”
§ Forecasting Stock Market Prices Using Machine Learning and Deep Learning Models: A Systematic Review, Performance Analysis and Discussion of Implications > 2. Comprehensive Summary of Theoretical Basis
What the sources say
- Supportedcites Qiu et al. 2020 source
The 'Long short-term memory neural networks (LSTM)' section describes LSTM as a form of RNN intended to avoid long-term dependence problems. The Introduction also says LSTM is suitable for stock-price time series. Page numbers are not provided in our copy. (LLM, conf=0.92)
- Not checked — no full text of the sourcecites Banik et al. 2022 source
plainstates a cause- 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.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“However, when it comes to predicting financial distress, the logistic regression model has exhibited better results in comparison to neural networks (Zizi et al. 2021).”
§ Forecasting Stock Market Prices Using Machine Learning and Deep Learning Models: A Systematic Review, Performance Analysis and Discussion of Implications > 2.3.1. Long Short-Term Memory (LSTM)
What the sources say
- Supportedcites Zizi et al. 2021 source
The Discussion reports that, in the authors’ comparison, logistic regression performed better than neural networks. The Conclusions likewise describe superior test-sample accuracy for logistic regression in predicting financial distress among Moroccan SMEs; the claim is supported as a report of those results, not as a universal rule. (LLM, conf=0.94)
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.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“ARIMA can be combined with another algorithm, symmetric generalized autoregressive conditional heteroskedasticity (SGARCH), to improve forecasting performance.”
§ Forecasting Stock Market Prices Using Machine Learning and Deep Learning Models: A Systematic Review, Performance Analysis and Discussion of Implications > 2.2.1. ARIMA
What the sources say
- Supportedcites Vo and Ślepaczuk 2022 source
The Methodology section describes a hybrid model that combines ARIMA with GARCH to forecast the mean and variance. The Results section reports that the ARIMA–SGARCH model predicted S&P 500 returns more accurately than ARIMA alone on the study’s error metrics. (LLM, conf=0.94)
plainstates a cause- 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.
- Supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Models with an integrated GNB algorithm will yield feature extraction and feature scaling outcomes that are superior to those already achieved using models that incorporate either the GNB algorithm or feature scaling (Setiani et al. 2020;Ampomah et al. 2021).”
§ Forecasting Stock Market Prices Using Machine Learning and Deep Learning Models: A Systematic Review, Performance Analysis and Discussion of Implications > Sr No Algorithm Name Gap Analysis Performance Evaluation 1 Linear regression
What the sources say
- Could not verifycites Setiani et al. 2020 sourceExperimental — accuracy not yet measured
The Methodology section describes a Naïve Bayes classifier and a Gaussian-distribution calculation, while Figure 3 reports 57% prediction accuracy. Neither passage addresses integrated GNB with feature extraction or scaling, or compares it with models using either component alone (searched: integrated GNB, feature extraction, feature scaling, comparative performance). (LLM, conf=0.76)
- Partly supportedcites Ampomah et al. 2021 sourceExperimental — accuracy not yet measured
The Conclusion reports that models combining GNB, feature scaling, and feature extraction generally outperformed models combining GNB with just one of those techniques. The cited paper measures stock-price prediction performance, not the quality of feature extraction or scaling themselves, and expressly notes an exception: GNB_LDA. (LLM, conf=0.86)
hedgedstates a cause- 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.
- Could not verify
7 more claims · 1 with their cited sources checked
Create a free account to see all 87 claims and to verify the 1 reference still open.
- Probably not a substantial claim. 50
- It may state two separate claims as one. 45
- The quote only partly states this claim. 2
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 major: it would touch a result — a value a result rests on, reported and recomputed p-values on opposite sides of the significance level, two passages that conflict with no sign which is right, or a cited work retracted before this paper was published.
In the paper
§ Forecasting Stock Market Prices Using Machine Learning and Deep Learning Models: A Systematic Review, Performance Analysis and Discussion of Implications > 7. Conclusions
“| Random forest | 87.8839 | 0.9818 | 98.5633 | 0.9428 |”
§ Forecasting Stock Market Prices Using Machine Learning and Deep Learning Models: A Systematic Review, Performance Analysis and Discussion of Implications > 7. Conclusions
“| SVR | 4.525 | 0.9279 | 1.5074 | 0.9432 |”
Rated major: it would touch a result — a value a result rests on, reported and recomputed p-values on opposite sides of the significance level, two passages that conflict with no sign which is right, or a cited work retracted before this paper was published.
In the paper
§ Forecasting Stock Market Prices Using Machine Learning and Deep Learning Models: A Systematic Review, Performance Analysis and Discussion of Implications > 7. Conclusions
“| SVR | 4.525 | 0.9279 | 1.5074 | 0.9432 |”
§ Forecasting Stock Market Prices Using Machine Learning and Deep Learning Models: A Systematic Review, Performance Analysis and Discussion of Implications > 7. Conclusions
“| MLPR | 2.5893 | 0.9611 | 2.4764 | 0.9472 |”
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
§ Forecasting Stock Market Prices Using Machine Learning and Deep Learning Models: A Systematic Review, Performance Analysis and Discussion of Implications > 2.2.2. FB Prophet Model
“where l(t) represents the linear trend, sp(t) represents seasonal patterns, v(t) represents holiday effects, and εn is the white noise error.”
§ Forecasting Stock Market Prices Using Machine Learning and Deep Learning Models: A Systematic Review, Performance Analysis and Discussion of Implications > 7 FB Prophet
“One of its advantages is that it does not consider holidays or null values.”
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
§ Forecasting Stock Market Prices Using Machine Learning and Deep Learning Models: A Systematic Review, Performance Analysis and Discussion of Implications > 3. General Machine Learning Pipeline
“Training the ML model involves six steps, which are described in the following section:”
§ Forecasting Stock Market Prices Using Machine Learning and Deep Learning Models: A Systematic Review, Performance Analysis and Discussion of Implications > 3. General Machine Learning Pipeline
“Step 7: Fine-tune the hyperparameters to improve the evaluation parameters of the model.”
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
§ Forecasting Stock Market Prices Using Machine Learning and Deep Learning Models: A Systematic Review, Performance Analysis and Discussion of Implications > 2. Comprehensive Summary of Theoretical Basis
“The model showed the lowest RMSE value of 186.32, a precision of 60%, and an F1-score of 66.47 (Li and Pan 2021).”
§ Forecasting Stock Market Prices Using Machine Learning and Deep Learning Models: A Systematic Review, Performance Analysis and Discussion of Implications > 13 XG-Boost + LSTM
“The blending ensemble algorithm, which consists of a modified version of RNN, i.e., LSTM and GRU, has the following evaluation parameters: MSE = 186.32,MPA = 99.65,precision = 60%, Recall = 75%, F1-Score = 66.67% (Li and Pan 2021).”
8 more findings
Create a free account to see every finding of this check.
Cited works that were retracted
Works this paper cites that were later retracted, from the retraction notices publishers deposit. Major when the notice came out before this paper was published.
A note, not an issue: for example, a cited work that was retracted only after this paper was published, which its authors could not have known.
In the paper
“In (Sharaf et al. 2022), news headlines pertaining to TSLA, AMZ, and GOOG stock were considered to obtain good-quality data to reduce spam tweets through feature selection methods.”
Cited work 10.1007/s11042-022-14216-w · retracted 2026-04-25 · notice 10.1007/s11042-026-21614-x · this paper published 2023-07-26
Citations
Cites
1 retracted
Sharaf, M. M., Hemdan, E. E., El‐Sayed, A., & El-Bahnasawy, N. A. W. (2022) · Multimedia Tools and Applications · doi:10.1007/s11042-022-14216-w · Cited by 65 · cited 1×
“In (Sharaf et al. 2022), news headlines pertaining to TSLA, AMZ, and GOOG stock were considered to obtain good-quality data to reduce spam tweets through feature selection methods.”
BackgroundRetractedPatel, J., Shah, S., Thakkar, P., & Kotecha, K. (2015) · Expert Systems with Applications · doi:10.1016/j.eswa.2014.07.040 · Cited by 1,141 · cited 1×
“Random forest, trained on multiple decision trees and Naïve Bayes, predicts stock movement based on negative or positive probability using Reliance and Infosys’ 10-year historical data (Patel et al. 2015).”
MethodSentiment analysis using deep learning architectures: a review
Yadav, A., & Vishwakarma, D. K. (2020) · Artificial Intelligence Review · doi:10.1007/s10462-019-09794-5 · Cited by 841 · cited 1×
“Social media data, company news, and trend analysis can classify investors’ stock sentiments as positive, negative, or neutral (Yadav and Vishwakarma 2019).”
BackgroundApplications of deep learning in stock market prediction: Recent progress
Jiang, W. (2021) · Expert Systems with Applications · doi:10.1016/j.eswa.2021.115537 · Cited by 714 · cited 2×
“We believe that researchers can explore the possible usage of state-of-the-art deep learning and machine learning algorithms in stock trend prediction and classification (Jiang 2021; Nikou et al. 2019).”
BackgroundEvaluating 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 · cited 7×
“The performance of the SVM algorithm can be elevated by tuning parameters such as regularization, gamma, and kernel parameters (Bustos et al. 2017).”
MethodInfluentialStock Market Analysis: A Review and Taxonomy of Prediction Techniques
Shah, D., Isah, H., & Zulkernine, F. (2019) · International Journal of Financial Studies · doi:10.3390/ijfs7020026 · Cited by 465 · cited 2×
“As the stock market evolves under the influence of various factors such as geopolitical issues, equity trading, and investment, the underlying challenges also change, and so do the methodologies for addressing the new challenges (Sprenger and Welpe 2011; Shah et al. 2019).”
BackgroundNabipour, M., Nayyeri, P., Jabani, H., S., S., & Mosavi, A. (2020) · IEEE Access · doi:10.1109/access.2020.3015966 · Cited by 463 · cited 1×
“Similar feature engineering-and deep learning-based approaches are presented in (Shen and Shafiq 2020; Nabipour et al. 2020; Mohapatra et al. 2022).”
BackgroundThe evolution of stock market efficiency over time: A survey of the empirical literature
Lim, K.-P., & Brooks, R. (2011) · Journal of Economic Surveys · doi:10.1111/j.1467-6419.2009.00611.x · Cited by 410 · cited 2×
“It is said that the stock market evolves over a period of time (Lim and Brooks 2011), and hence, the approaches developed for handling specific problems will see low performance sooner or later even though their performance is found to be appreciated initially.”
BackgroundShort-term stock market price trend prediction using a comprehensive deep learning system
Shen, J., & Shafiq, M. O. (2020) · Journal of Big Data · doi:10.1186/s40537-020-00333-6 · Cited by 380 · cited 1×
“Similar feature engineering-and deep learning-based approaches are presented in (Shen and Shafiq 2020; Nabipour et al. 2020; Mohapatra et al. 2022).”
BackgroundForecasting stock prices with long-short term memory neural network based on attention mechanism
Qiu, J., Wang, B., & Zhou, C. (2020) · PLOS ONE · doi:10.1371/journal.pone.0227222 · Cited by 366
Research works printed but not linked: 4
- Baheti, 2021, Stock Market Analysis from Social Media and News using Machine Learning Techniques, International Journal on Data Science and Machine Learning with Applications, 1, 59
- Gururaj, 2019, Stock Market Prediction using Linear Regression and Support Vector Machines, International Journal of Applied Engineering Research, 14, 1931
- Seethalakshmi, 2018, Analayis of stock market predictor variables using linear regression, International Journal of Pure and Applied Mathematics, 119, 369
- Venkat, 2022, Stock Market Trend Prediction Using K-Nearest Neighbor (KNN) Algorithm, Journal of Engineering Sciences, 3, 32
Cited by
Shaban, W. M., Ashraf, E., & Slama, A. E. (2023) · Neural Computing and Applications · doi:10.1007/s00521-023-09179-4 · Cited by 124
“These methods are ABC-ANFIS-SVM [16], DLSMP [17], CNN-BiLSTM-AM [20], PSO-LSTM [22], BiLSTM [26], CNN [27], and CKELM [28] as shown in Table 1.”
Rane, N. L., Choudhary, S., & Rane, J. (2023) · SSRN Electronic Journal · doi:10.2139/ssrn.4640316 · Cited by 141
Deep Learning in Finance: A Survey of Applications and Techniques
Mienye, E. D., Jere, N. R., Obaido, G., Mienye, I. D., & Aruleba, K. (2024) · AI · doi:10.3390/ai5040101 · Cited by 114
Deep learning for financial forecasting: A review of recent trends
Giantsidi, S., & Tarantola, C. (2025) · International Review of Economics & Finance · doi:10.1016/j.iref.2025.104719 · Cited by 104
Vuong, P. H., Phu, L. H., Nguyen, T. H. V., Duy, L. N., Bao, P. T., & Trinh, T. D. (2024) · Science Progress · doi:10.1177/00368504241236557 · Cited by 57
“1 Numerous models and predictions of stock prices have been proposed.”
Background
Metadata
- Authors
- Sonkavde, G., Dharrao, D. S., Bongale, A. M., Deokate, S. T., Doreswamy, D., & Bhat, S. K.
- Affiliations
- Symbiosis International University (IN)
- Savitribai Phule Pune University (IN)
- Department of Mechatronics, Manipal Institute of Technology, Manipal Academy of Higher Education, Manipal 576104, Karnataka, India
- Manipal Academy of Higher Education (IN)
- Corresponding author
- Dharrao, Deepak Sudhakar
- Bongale, Anupkumar M.
- ORCID
- Venue
- International Journal of Financial Studies
- Publisher
- MDPI AG
- Published
- 2023-07-26
- Volume
- 11
- Issue
- 3
- Pages
- 94
- ISSN
- 2227-7072 (ISSN-L)
- OpenAlex
- W4385363975
- Cited by
- 262
- References
- 83
- Citations by year
- 2026: 72 · 2025: 125 · 2024: 59 · 2023: 6
- FWCI
- 8.55
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- International Journal of Financial Studies (publishedVersion, cc-by)
- RePEc: Research Papers in Economics (submittedVersion)
- 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 › Economics, Econometrics and Finance › Finance › Financial Markets and Investment Strategies
- Social Sciences › Business, Management and Accounting › Accounting › Financial Analysis and Corporate Governance
- Keywords
- stock price prediction, financial time series forecasting, ensemble learning, random forest, XGBoost, LSTM, financial classification, artificial intelligence in finance, portfolio optimization
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 2026-09-28, under version 3 of the claim definition (prompt extract-v3).
A second model labelled each claim on 2026-09-28 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
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: Forecasting Stock Market Prices Using Machine Learning and Deep Learning Models: A Systematic Review, Performance Analysis and Discussion of Implications. https://app.cochat.ai/papers/10.3390/ijfs11030094
- The paper
- Sonkavde, G., Dharrao, D. S., Bongale, A. M., Deokate, S. T., Doreswamy, D., & Bhat, S. K. (2023). Forecasting Stock Market Prices Using Machine Learning and Deep Learning Models: A Systematic Review, Performance Analysis and Discussion of Implications. International Journal of Financial Studies. https://doi.org/10.3390/ijfs11030094
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.