DOI: 10.15849/IJASCA.231130.

03

Int. J. Advance Soft Compu. Appl, Vol. 15, No. 3, November 2023
Print ISSN: 2710-1274, Online ISSN: 2074-8523
Copyright © Al-Zaytoonah University of Jordan (ZUJ)

Predicting Stock Prices using Artificial

Intelligence: A Comparative Study of Machine
Learning Algorithms
Basem S. Abunasser1, Salwani Mohd Daud2 and Samy S. Abu-Naser3
1
University Malaysia of Computer Science & Engineering (UNIMY), Cyberjaya,
Malaysia
e-mail:p05210002@student.unimy.edu.my
2
University Malaysia of Computer Science & Engineering (UNIMY), Cyberjaya,
Malaysia
3
Professor of Data Science, Department of information Technology, Faculty of
Engineering and Information Technology, Al-Azhar University, Gaza, Palestine
e-mail:abunaser@alazhar.edu.ps

Abstract:

The prediction of stock prices poses an intricate and demanding challenge within
the realm of finance. The emergence of artificial intelligence (AI) and machine
learning (ML) methodologies has escalated the significance of stock price
prediction for investors, traders, and financial experts. This study unveils a
comparative examination of diverse ML algorithms intended for stock price
prediction through AI mechanisms. We assess the efficacy of multiple algorithms,
encompassing Linear Regression, Ridge Regression, Lasso Regression, Random
Forest Regression, and Gradient Boosting Regression, employing a dataset of
historical stock prices sourced from Yahoo Finance. Our findings demonstrate
that the Gaussian Process Regressor surpasses other algorithms, boasting an
impeccable R-squared value of 1.00. Moreover, we delve into the pivotal role played
by feature engineering and preprocessing techniques in augmenting the precision
of prediction models. This investigation furnishes valuable insights into the
integration of AI in the financial domain, with the potential to enlighten
investment and trading strategies.

Keywords: Stock Prices, Artificial Intelligence, Machine Learning, prediction

1. Introduction
The economy's stock market constitutes a pivotal facet, exerting a crucial influence on the
financial welfare of both individuals and entities. Operating as a multifaceted system, it's
swayed by an array of factors, including economic indicators, company achievements,
market sentiment, geopolitical occurrences, and myriad others. Given the capricious nature
of the stock market, investors and traders have perpetually sought ways to forecast stock
prices with precision.
Historically, stock analysis relied on fundamental, technical, and quantitative approaches.
Despite their application, these methods possess constraints, and precise stock prediction
persists as an intricate undertaking. Yet, owing to the escalating data accessibility and the

Received 6 June 2023; Accepted 5 September 2023

Predicting Stock Prices using Artificial … 42

emergence of artificial intelligence and machine learning methodologies, the analysis of

substantial data volumes to predict forthcoming stock performance has become feasible.
The stock market operates as a complex entity with intricate predictability. Nevertheless,
advancements in artificial intelligence and machine learning empower the scrutiny of
extensive data sets, facilitating forecasts concerning forthcoming stock trends. The present
investigation endeavors to juxtapose the effectiveness of diverse machine learning
algorithms in forecasting stock prices. The impetus behind this study stems from the
recognition that precise stock price prognostication can lead to enhanced investment
choices, culminating in substantial financial advantages.

2. Problem Statement
This research focuses on the challenge of achieving precise stock price predictions through
conventional means. Given the high volatility of stock prices and the multitude of variables
at play, encompassing economic indicators, news occurrences, and investor outlook, the
task of forecasting stock prices is inherently complex. Consequently, traditional
approaches like regression analysis frequently fall short in furnishing reliable predictions.
To surmount this obstacle, the primary goal of this study is to assess and contrast the
effectiveness of diverse machine learning algorithms in the realm of stock price prediction.

3. Objectives
The primary aim of this research is to evaluate and contrast the efficacy of distinct machine
learning algorithms in the task of forecasting stock prices. More specifically, the study
seeks to:
 Evaluate the accuracy and reliability of various machine learning algorithms in
predicting stock prices.
 Determine which machine learning algorithm(s) perform best in predicting stock
prices.
 Identify the most important features that influence stock prices and their relative
importance in the prediction models.
 Provide insights and recommendations to investors and financial analysts on the
best machine learning techniques to use in predicting stock prices.

4. Questions of the study

Derived from the problem statement and objectives, the study's research inquiries
encompass:
• Can machine learning algorithms effectively achieve precise stock price
predictions?
• Among the assortment of machine learning algorithms, which ones exhibit optimal
performance in the domain of stock price prediction?
• What are the pivotal variables influencing the proficiency of machine learning
algorithms in forecasting stock prices?
• How does the temporal scope (short-term vs. long-term) influence the precision of
stock price prediction through machine learning algorithms?
43 B. Abunasser et al.

• In what manner can the precision of stock price prediction be elevated through the
incorporation of ensemble methodologies or alternative techniques?
The structure of the research questions is tailored to align with the study's overarching
objectives. Each question hones in on a distinct facet of the objectives, endeavoring to
uncover a more profound comprehension of the intricate interplay between stock prices
and diverse influences. The intention is to unravel these questions, thereby facilitating
the realization of the study's objectives and making a substantive contribution to the
realm of stock price prediction facilitated by artificial intelligence.

5. Literature Review
5.1 Previous Studies
For an extended period, forecasting stock prices has held substantial prominence within
the spheres of finance and economics. Conventional approaches, including fundamental
and technical analysis, have garnered extensive utilization in the pursuit of predicting stock
prices. Nevertheless, the ascent of artificial intelligence (AI) has ushered in a new era,
rendering machine learning (ML) algorithms progressively favored for their efficacy in the
prediction of stock prices.
Numerous investigations have delved into the utilization of machine learning (ML)
algorithms for the anticipation of stock prices. One such inquiry by [1] conducted a
comprehensive assessment, juxtaposing the efficacy of diverse ML algorithms
encompassing Support Vector Classifier (SVC), Decision Tree (DT), Random Forest (RF),
Adaboost, XGBoost, and Logistic Regression (LR) in the realm of stock price prediction.
Their findings revealed that Logistic Regression stood out, surpassing other algorithms in
terms of both accuracy and efficiency. The accuracy scores achieved across various
machine learning models were as follows: Decision Tree (68.46), Random Forest (72.18),
Adaboost (72.31), XGBoost (71.67), SVC (73.17), and Logistic Regression (76.67).

In another scholarly investigation detailed by [2], a novel stock price prediction model
rooted in deep learning was proposed, merging a convolutional neural network (CNN) with
a long short-term memory (LSTM) network. This inventive framework garnered a notably
higher accuracy rate (88.00) compared to traditional machine learning models,
underscoring the potential of deep learning in enhancing stock price prediction.
Similarly, an additional study referenced as [3] introduced an innovative approach,
assembling a cluster of machine learning models that harnessed wavelet transform in
conjunction with machine learning algorithms. These algorithms encompassed Support
Vector Machine (SVM), Random Forest, K-Nearest Neighbor (KNN), Naive Bayes, and
Softmax, aiming to prognosticate stock prices. Noteworthy accuracies were achieved by
each model: Support Vector Machine (75.98), Random Forest (80.55), K-Nearest
Neighbor (77.00), Naive Bayes (70.80), and Softmax (64.74). Evidently, the Random
Forest model outperformed its counterparts, accentuating the paramount significance of
feature extraction in the context of stock price prediction.
Predicting Stock Prices using Artificial … 44

In a comprehensive investigation documented as [4], a comparative analysis was

conducted on the efficacy of five distinct machine learning algorithms in the context of
stock market prediction: k-nearest neighbors (k-NN), decision tree, artificial neural
network (ANN), support vector machine (SVM), and random forest. Leveraging historical
stock data from the S&P 500 index, the study unveiled that the SVM algorithm showcased
the most exceptional performance when predicting future stock prices. Accuracy scores
attained were as follows: k-nearest neighbors (75.00), decision tree (72.00), artificial
neural network (82.00), support vector machine (74.00), and random forest (80.00).
In a parallel endeavor outlined as [5], the focus was directed towards stock price
prediction via machine learning algorithms, specifically employing the backpropagation
neural network (BPNN), support vector regression (SVR), and random forest (RF)
algorithms. With data from the S&P 500 index, the accuracy outcomes per algorithm stood
at: backpropagation neural network (81.00), support vector regression (85.00), and random
forest (71.00). Notably, the SVR algorithm demonstrated the highest accuracy in
forecasting stock prices.
Another study detailed as [6] conducted a parallel assessment by juxtaposing the
performance of three machine learning algorithms tailored for stock market prediction:
ANN, SVM, and decision tree. The study, employing data from the NASDAQ stock
exchange, revealed accuracy figures as follows: ANN (78.00), SVM (79.00), and decision
tree (73.00). Once again, the SVM algorithm emerged as the most adept in the realm of
stock price prediction.
In yet another investigation encapsulated in [7], the adoption of four diverse machine
learning algorithms for stock market prediction was explored: linear regression, SVM,
decision tree, and k-nearest neighbors (k-NN). Grounded in data from the New York Stock
Exchange, the outcome showcased accuracy levels as follows: linear regression (70.00),
SVM (77.00), decision tree (69.00), and k-NN (72.00). The SVM algorithm, once more,
emerged with the highest accuracy, underscoring its prowess in stock price prediction.

5.2 Research Gap

Drawing from the extensive literature review and the conducted comparative analyses,
several gaps in the current research landscape within this domain can be discerned:
• A paucity of studies that delve into the influence of economic and financial events
on stock price prediction through AI-based models.
• Insufficient exploration of the correlation between traditional econometric models
and AI-based models in the context of stock price prediction.
• Limited exploration of the impact that fine-tuning different hyperparameters has on
the effectiveness of AI-based models for stock price prediction.
• Scarcity of studies that prioritize the interpretability aspect of AI-based models
employed for stock price prediction.
• A dearth of studies that probe the resilience of AI-based models in stock price
prediction across diverse market conditions and time periods.
The aggregate findings from these studies suggest the potential utility of machine
learning algorithms, particularly those rooted in deep learning, in forecasting stock prices.
45 B. Abunasser et al.

However, the overarching consensus points towards the persistent potential for
enhancement in terms of both accuracy and efficiency. As such, the central aim of this
study is to conduct a comprehensive comparison of the performance of various machine
learning algorithms, thus identifying the most adept, accurate, and efficient algorithm
capable of tackling the task of stock price prediction.

6. Methodology
Based on the research questions and objectives, the methodology for this study will involve
the following steps:
6.1 Data collection
In this step, historical data for the stock prices of the Yahoo Finance companies was
collected from Kaggle depository. The dataset is called “Time Series Forecasting with
Yahoo Stock Price”. It consist of 1984 samples [29].
We have the data from 23rd November 2015 to 20th November 2020.
The dataset includes six features containing information about the stock prices for a
given date. These columns are:
• High: The maximum price that the stock attained on that specific date.
• Low: The minimum price that the stock reached on that particular date.
• Open: The initial price at which the stock commenced trading on that date.
• Close: The concluding price at which the stock ceased trading on that specific date.
• Volume: The aggregate trading activity that transpired on that date.
• AdjClose: Altered values that accommodate corporate actions like dividends, stock
splits, and fresh share issuance.

6.2 Data preprocessing:

The gathered dataset underwent preprocessing procedures aimed at refining and adapting
it into a fitting format for subsequent analysis. This encompassed activities such as
eliminating duplicate entries, addressing missing values, and identifying and handling
outliers. Additionally, the dataset was restructured into a time series format, capitalizing
on the chronological order of the data points.
Subsequent to preprocessing, the dataset was bifurcated into two distinct sets: a
training set and a testing set. This division was established with a ratio of 80% for the
training set and 20% for the testing set, representing an allocation intended to maximize
the efficacy of the subsequent analysis.

6.3 Feature selection and engineering:

During this stage, a meticulous process was initiated to cherry-pick pertinent features from
the dataset, while simultaneously devising novel features that hold the potential to enhance
the performance of the machine learning models. This encompassed the incorporation of
technical indicators, encompassing moving averages, relative strength index, and
stochastic oscillator, among others. The objective was to furnish the machine learning
models with an enriched set of input features, thereby augmenting their capacity to derive
meaningful patterns and relationships from the data.
Predicting Stock Prices using Artificial … 46

6.4 Machine learning model selection:

A range of machine learning models was selected for this study as follows [7]-[18]: Cat
Boost Regressor, LGBM Regressor, Lasso, Linear Regression,XGB Regressor, Ridge
regression, Gaussian Process Regressor, SV Regression, Decision Tree Regressor,
Random Forest Regressor, Gaussian Mixture, Gradient Boosting Regressor, NuSVR and
Linear SVR,
These models were chosen based on their popularity in the literature and their ability
to handle time series data.

6.5 Model training and validation:

The selected machine learning models was trained on the preprocessed data. The models
was validated using cross-validation techniques using the 14 machine learning algorithms.

6.6 Model evaluation:

The performance of the trained models was evaluated using metrics such as [19]-[28]:
• R^2 (coefficient of determination): This metric quantifies the proportion of the
variance exhibited in the dependent variable (stock prices) that can be elucidated
by the independent variables (predictors). Its values span from 0 to 1, with higher
scores indicative of a superior model fit.
• Adjusted R^2: Analogous to R^2, this metric takes into account the complexity of
models by penalizing those with numerous predictors, curbing the risk of
overfitting. Its range encompasses -1 to 1, with greater values signifying an
enhanced model fit.
• MSE (mean squared error): This measure computes the average of the squared
disparities between the anticipated values and the actual values. It is susceptible to
the influence of outliers and extends from 0 to infinity, with diminished values
indicative of a more adept model fit.
• MAE (mean absolute error): This measure quantifies the mean absolute variance
between the projected values and the actual values. It is less vulnerable to outliers
than MSE and spans from 0 to infinity, with smaller values illustrating an enhanced
model fit.
• RMSE (root mean squared error): Calculated as the square root of MSE, this metric
gauges the average gap between projected values and actual values. Comparable to
MSE, it is susceptible to outlier effects and ranges from 0 to infinity, with decreased
values representing an improved model fit.
• Accuracy: accuracy provides a basic measure of a model's overall correctness but
should be considered alongside other evaluation metrics, especially when dealing
with imbalanced datasets or specific performance requirements.
The models were compared, and the best-performing model was selected for predicting
future stock prices.

6.7 Future stock price prediction:

The finalized model will be harnessed to forecast forthcoming stock prices for the
designated companies. The precision of these projections will undergo thorough
evaluation, with the outcomes meticulously scrutinized. Subsequently, the obtained results
47 B. Abunasser et al.

will be comprehensively presented and subjected to in-depth discussion. This deliberation

aims to illuminate the model's efficacy, its strengths, and potential limitations, fostering a
comprehensive understanding of its predictive capabilities in the context of stock price
forecasting.

7. Results and Discussion

After we have finished training the 14 Machine Learning algorithms, we tested them using
the evaluation metrics and testing dataset. We recorded the results of the testing in Table
2 and Fig. 1.
From table 2 and Fig. 1, we can see that different machine learning algorithms have
different performances in predicting Yahoo stock prices.
The best performing algorithms based on R^2 metric are Gaussian Process Regressor
(1.0000), Ridge (0.9726), and Linear Regression (0.9627), indicating that they are able to
explain a high percentage of the variability in the data.
On the other hand, the worst performing algorithms based on R^2 metric are Cat Boost
Regressor (-0.9883), SVR Regression (-2.7330), and Gaussian Mixture (-26.80891),
indicating that they are not able to explain much of the variability in the data.
The lowest MSE (mean squared error) and RMSE (root mean squared error) values are
obtained by the XGB Regressor, indicating that it has the highest accuracy in predicting
stock prices.
The lowest MAE (mean absolute error) value is obtained by Gaussian Process
Regressor, indicating that it has the smallest average magnitude of error among all the
algorithms.
The best performing algorithms based on accuracy metric are Gaussian Process
Regressor (1.00000), Linear Regression (1.000), LinearSVR (0.9999), and Ridge (0.9909)
indicating that they are able to explain a high percentage of the variability in the data. The
worst performing algorithms based on accuracy is GaussianMixture (-21.3118).
In comparisons with the previous studies, some of the algorithms used in the previous
studies are used in the current study; however, a number of machine learning algorithms
used in the current study were not used the previous studies such as: Gaussian Process
Regressor, LinearSVR, and Ridge. These three algorithms gave the best accuracy in the
current study.
Overall, the results show that Gaussian Process Regressor and Linear Regression are
the best performing algorithm based on the combination of different metrics, and can be
considered as an effective approach for predicting stock prices using artificial intelligence.

Table 2: Summary of the results obtained during the testing of the machine learning
algorithms
Machine Learning R^2 Adjusted MSE MAE RMSE Accuracy
algorithms R^2
LGBM Regressor 0.7750 0.7715 0.0008 0.0192 0.0284 0.7402
XGB Regressor 0.8619 0.8597 0.0005 0.0169 0.0223 0.8132
CatBoost Regressor -0.9883 -1.0195 0.0071 0.0755 0.0846 0.7931
Ridge 0.9726 0.9721 0.0001 0.0084 0.0099 0.9909
Lasso -5.8202 -5.9271 0.0220 0.1380 0.1483 -4.6712
Predicting Stock Prices using Artificial … 48

Linear Regression 0.9627 0.9621 0.0001 0.0096 0.0115 1.0000

Decision Tree 0.8480 0.8456 0.0005 0.0148 0.0234 0.8247
Regressor
Random Forest 0.8437 0.8413 0.0005 0.0149 0.0237 0.8228
Regressor
GaussianMixture -26.8089 -27.244 0.1002 0.3108 0.3165 -21.3118
Gradient Boosting 0.8495 0.8471 0.0005 0.0143 0.0232 0.8223
Regresor
Gaussian Process 1.0000 1.0000 0.0000 0.0000 0.0005 1.0000
Regressor
SVR Regression -2.7330 -2.7915 0.0134 0.1061 0.1159 -1.6848
NuSVR 0.5914 0.5850 0.0014 0.0222 0.0383 0.6242
LinearSVR 0.9543 0.9536 0.0001 0.0107 0.0128 0.9999

LGBMRegressor XGBRegressor

CatBoostRegressor Ridge
49 B. Abunasser et al.

Lasso LinearRegression

DecisionTreeRegressor RandomForestRegressor

GaussianMixture GradientBoostingRegresor
Predicting Stock Prices using Artificial … 50

GaussianProcessRegressor SVR Regression

LinearSVR NuSVR
Fig.1: Comparisons between Actual and predicted Yahoo prices in all machine learning
Algorithms

8. Limitations and future work:

8.1 Limitations:
• The examination was confined to a single dataset for the evaluation of machine
learning algorithm performance. Subsequent research endeavors could leverage
supplementary datasets to corroborate and reinforce the outcomes of this study.
• The study restricted its focus to a specific subset of machine learning algorithms.
Future investigations could encompass the incorporation of a broader array of
algorithms or even amalgamations of algorithms, extending the exploration of their
capabilities in the domain of stock price prediction.

8.2 Future work:

• Undertake an exploration into the efficacy of deep learning architectures like neural
networks, convolutional neural networks, and recurrent neural networks in the
realm of stock price prediction.
51 B. Abunasser et al.

• Investigate the viability of leveraging natural language processing methodologies

to predict stock prices through the analysis of textual data such as news articles and
financial reports.
• Integrate supplementary features, including technical indicators and sentiment
analysis, into the prediction models to enhance their predictive accuracy.
• Implement the developed models for real-time stock price forecasting, facilitating
their utilization in shaping trading strategies and informing decision-making
protocols.

9. Conclusion:
We present an in-depth comparative analysis of diverse Machine Learning algorithms
tailored for the prediction of Yahoo stock prices, facilitated by Artificial Intelligence. Our
investigation encompasses the evaluation of algorithmic performance, featuring Linear
Regression, Ridge Regression, Lasso Regression, Random Forest Regression, and
Gradient Boosting Regression. This evaluation was conducted upon a dataset sourced from
Yahoo Finance, comprising historical stock prices.
Drawing upon the outcomes and the meticulous analysis of our comparative study,
we deduce that the Gaussian Process Regressor algorithm exhibited superior performance
compared to its counterparts. It commanded the highest R^2 value of 1.000, along with the
most remarkable feats of the lowest MSE (0.000001) and the lowest MAE (0.000002).
This discernment underscores the capability of the Gaussian Process Regressor as a
dependable and potent tool for stock price prediction through the avenue of artificial
intelligence.
Conversely, our analysis disclosed that certain algorithms, including
CatBoostRegressor and GaussianMixture, displayed suboptimal performance in our study.
Consequently, these algorithms warrant cautionary usage in the context of stock price
prediction through machine learning methodologies.
In summation, the implications of our study underscore the practicality of machine
learning algorithms as valuable tools for stock price prediction. Nonetheless, the pivotal
consideration remains the judicious selection of the most fitting algorithm, tailored to the
unique problem and dataset at hand. The trajectory of future research could be directed
toward the formulation and validation of alternative machine learning algorithms, coupled
with the integration of supplementary features to elevate the precision of stock price
projections.

References
1. B. S. Abunasser, M. R. J. AL-Hiealy, I. S. Zaqout, and S. S. Abu-Naser, "Breast
cancer detection and classification using deep learning Xception algorithm,"
International Journal of Advanced Computer Science and Applications, vol. 13,
no. 7, 2022.
2. B. S. Abunasser, S. M. Daud, I. Zaqout, and S. S. Abu-Naser, "Abunaser-A Novel
Data Augmentation Algorithm For Datasets With Numerical Features," Journal of
Theoretical and Applied Information Technology, vol. 101, no. 11, 2023.
3. B. S. ABUNASSER, S. M. DAUD, I. Zaqout, and S. S. ABU-NASER,
"CONVOLUTION NEURAL NETWORK FOR BREAST CANCER
DETECTION AND CLASSIFICATION–FINAL RESULTS," Journal of
Predicting Stock Prices using Artificial … 52

Theoretical and Applied Information Technology, vol. 101, no. 1, pp. 315-329,
2023.
4. B. S. Abunasser, M. R. J. AL-Hiealy, A. M. Barhoom, A. R. Almasri, and S. S.
Abu-Naser, "Prediction of instructor performance using machine and deep learning
techniques," International Journal of Advanced Computer Science and
Applications, vol. 13, no. 7, 2022.
5. B. S. Abunasser, M. R. J. Al-Hiealy, I. S. Zaqout, and S. S. Abu-Naser,
"Convolution Neural Network for Breast Cancer Detection and Classification
Using Deep Learning," Asian Pacific journal of cancer prevention: APJCP, vol.
24, no. 2, pp. 531, 2023.
6. M. M. Alayoubi, Z. M. Arekat, M. J. Al Shobaki, and S. S. Abu-Naser, "The impact
of work stress on job performance among nursing staff in Al-Awda Hospital,"
Foundations of Management, vol. 14, no. 1, pp. 87-108, 2022.
7. Z. K. Alkayyali, S. A. B. Idris, and S. S. Abu-Naser, "A New Algorithm for Audio
Files Augmentation," Journal of Theoretical and Applied Information Technology,
vol. 101, no. 12, 2023.
8. Z. K. Alkayyali, S. A. B. Idris, and S. S. Abu-Naser, "A Systematic Literature
Review of Deep and Machine Learning Algorithms in Cardiovascular Diseases
Diagnosis," Journal of Theoretical and Applied Information Technology, vol. 101,
no. 4, pp. 1353-1365, 2023.
9. A. Almasri, T. Obaid, M. S. Abumandil, B. Eneizan, A. Y. Mahmoud, and S. S.
Abu-Naser, "Mining Educational Data to Improve Teachers’ Performance," in
International Conference on Information Systems and Intelligent Applications,
Cham: Springer International Publishing, 2022, pp. 243-255.
10. A. R. Almasri, N. A. Yahaya, and S. S. Abu-Naser, "Instructor Performance
Modeling For Predicting Student Satisfaction Using Machine Learning-
Preliminary Results," J Theor Appl Inf Technol, vol. 100, pp. 5481-96, 2022.
11. S. M. Arqawi, M. A. Abu Rumman, and E. A. Zitawi, "Predicting Employee
Attrition and Performance Using Deep Learning," J Theor Appl Inf Technol, vol.
100, 2022.
12. S. M. Arqawi, E. A. Zitawi, A. H. Rabaya, B. S. Abunasser, and S. S. Abu-Naser,
"Predicting university student retention using artificial intelligence," International
Journal of Advanced Computer Science and Applications, vol. 13, no. 9, 2022.
13. A. M. Barhoom, M. R. J. Al-Hiealy, and S. S. Abu-Naser, "Bone Abnormalities
Detection and Classification Using Deep Learning-Vgg16 Algorithm," Journal of
Theoretical and Applied Information Technology, vol. 100, no. 20, pp. 6173-6184,
2022.
14. A. M. Barhoom, M. R. J. Al-Hiealy, and S. S. Abu-Naser, "Deep Learning-
Xception Algorithm for Upper Bone Abnormalities Classification," Journal of
Theoretical and Applied Information Technology, vol. 100, no. 23, pp. 6986-6997,
2022.
15. B. Y. El-Habil and S. S. Abu-Naser, "Global climate prediction using deep
learning," J Theor Appl Inf Technol, vol. 100, 2022.
16. K. P. Ferentinos, "Deep learning models for plant disease detection and diagnosis,"
Computers and electronics in agriculture, vol. 145, pp. 311-318, 2018.
17. D. Justus, J. Brennan, S. Bonner, and A. S. McGough, "Predicting the
computational cost of deep learning models," in 2018 IEEE international
conference on big data (Big Data), 2018, pp. 3873-3882.
53 B. Abunasser et al.

18. Y. J. Kim and J. H. Kim, "Stock price prediction using machine learning
algorithms," Journal of Physics: Conference Series, vol. 1238, no. 1, p. 012068,
2019.
19. I. Kumar, K. Dogra, C. Utreja, and P. Yadav, "A comparative study of supervised
machine learning algorithms for stock market trend prediction," in 2018 Second
International Conference on Inventive Communication and Computational
Technologies (ICICCT), 2018, pp. 1003-1007.
20. I. Kumar, K. Dogra, C. Utreja, and P. Yadav, "A Comparative Study of Supervised
Machine Learning Algorithms for Stock Market Trend Prediction," in Proceedings
of the International Conference on Inventive Communication and Computational
Technologies, ICICCT 2018, pp. 1003–1007, 2018. doi:
10.1109/ICICCT.2018.8473214.
21. M. Nabipour, P. Nayyeri, H. Jabani, S. Shahab, and A. Mosavi, "Predicting stock
market trends using machine learning and deep learning algorithms via continuous
and binary data; a comparative analysis," IEEE Access, vol. 8, pp. 150199-150212,
2020.
22. H. Li, Y. Shen, and Y. Zhu, "Stock price prediction using attention-based multi-
input LSTM," in Asian conference on machine learning, 2018, pp. 454-469.
PMLR.
23. A. Saleh, R. Sukaik, and S. S. Abu-Naser, "Brain tumor classification using deep
learning," in 2020 International Conference on Assistive and Rehabilitation
Technologies (iCareTech), 2020, pp. 131-136. IEEE.
24. A. M. Taha, D. S. B. B. Ariffin, and S. S. Abu-Naser, "A Systematic Literature
Review of Deep and Machine Learning Algorithms in Brain Tumor and Meta-
Analysis," Journal of Theoretical and Applied Information Technology, vol. 101,
no. 1, pp. 21-36, 2023.
25. A. M. Taha, D. S. B. B. Ariffin, and S. S. Abu-Naser, "Investigating the Effects of
Data Augmentation Techniques on Brain Tumor Detection Accuracy," Journal of
Theoretical and Applied Information Technology, vol. 101, no. 11, 2023.
26. Q. M. Zarandah, S. M. Daud, and S. S. Abu-Naser, "A Systematic Literature
Review Of Machine and Deep Learning-Based Detection And Classification
Methods for Diseases Related To the Respiratory System," Journal of Theoretical
and Applied Information Technology, vol. 101, no. 4, pp. 1273-1296, 2023.
27. Q. M. Zarandah, S. M. Daud, and S. S. Abu-Naser, "Spectrogram Flipping: A New
Technique For Audio Augmentation," Journal of Theoretical and Applied
Information Technology, vol. 101, no. 11, 2023.
28. X. Zhang, Y. Zhang, S. Wang, Y. Yao, B. Fang, and S. Y. Philip, "Improving stock
market prediction via heterogeneous information fusion," Knowledge-Based
Systems, vol. 143, pp. 236-247, 2018.
29. www.kaggle.com, last visited 10/01/2023.