The challenge for investors is the uncertainty of investing in a diversified portfolio using traditional techniques. This research project aimed to investigate and compare the effectiveness of two traditional Markowitz models and two machine learning techniques in optimizing diversified stock portfolios. The Markowitz models employed were the modern portfolio theory efficient frontier and quadratic mean-variance programming. The machine learning models utilized were the random forest classifier and the Extreme Gradient Boosting classifier. South African historical stock market time series listed companies on the Johannesburg Stock Exchange were analyzed to identify patterns and allocate stock weights to inform optimal stock portfolio diversification decisions. This minimizes risk while maximizing returns. Furthermore, the results are benchmarked against the JSE Top 40 Index. The portfolio generated by the eXtreme Gradient Boosting model performed the best, achieving a Sharpe ratio of 1.44. It is followed by the portfolio generated by the Random Forest classifier with a Sharpe ratio equal to 1.30. The portfolios generated by Modern Portfolio Theory Efficient Frontier and Quadratic Mean-Variance Programming had Sharpe ratios of 0.89 and 0.41, respectively. However, the JSE Top 40 Index portfolio had the lowest Sharpe ratio of 0.35. These results confirm the superiority of machine learning models over traditional Markowitz models. However, traditional models remain fundamental in understanding the principles of portfolio diversification. Machine learning models may be prioritized by investors primarily because of their ability to capture stochastic insights and market dynamics.
Citation: Esau Moyoweshumba, Modisane Seitshiro. Leveraging Markowitz, random forest, and XGBoost for optimal diversification of South African stock portfolios[J]. Data Science in Finance and Economics, 2025, 5(2): 205-233. doi: 10.3934/DSFE.2025010
Related Papers:
[1]
Hamdy M. Youssef, Najat A. Alghamdi, Magdy A. Ezzat, Alaa A. El-Bary, Ahmed M. Shawky .
A new dynamical modeling SEIR with global analysis applied to the real data of spreading COVID-19 in Saudi Arabia. Mathematical Biosciences and Engineering, 2020, 17(6): 7018-7044.
doi: 10.3934/mbe.2020362
[2]
H. M. Srivastava, I. Area, J. J. Nieto .
Power-series solution of compartmental epidemiological models. Mathematical Biosciences and Engineering, 2021, 18(4): 3274-3290.
doi: 10.3934/mbe.2021163
[3]
Hongfan Lu, Yuting Ding, Silin Gong, Shishi Wang .
Mathematical modeling and dynamic analysis of SIQR model with delay for pandemic COVID-19. Mathematical Biosciences and Engineering, 2021, 18(4): 3197-3214.
doi: 10.3934/mbe.2021159
[4]
Salma M. Al-Tuwairqi, Sara K. Al-Harbi .
Modeling the effect of random diagnoses on the spread of COVID-19 in Saudi Arabia. Mathematical Biosciences and Engineering, 2022, 19(10): 9792-9824.
doi: 10.3934/mbe.2022456
[5]
Tao Chen, Zhiming Li, Ge Zhang .
Analysis of a COVID-19 model with media coverage and limited resources. Mathematical Biosciences and Engineering, 2024, 21(4): 5283-5307.
doi: 10.3934/mbe.2024233
[6]
Sarafa A. Iyaniwura, Musa Rabiu, Jummy F. David, Jude D. Kong .
Assessing the impact of adherence to Non-pharmaceutical interventions and indirect transmission on the dynamics of COVID-19: a mathematical modelling study. Mathematical Biosciences and Engineering, 2021, 18(6): 8905-8932.
doi: 10.3934/mbe.2021439
[7]
A. Q. Khan, M. Tasneem, M. B. Almatrafi .
Discrete-time COVID-19 epidemic model with bifurcation and control. Mathematical Biosciences and Engineering, 2022, 19(2): 1944-1969.
doi: 10.3934/mbe.2022092
[8]
Cheng-Cheng Zhu, Jiang Zhu .
Spread trend of COVID-19 epidemic outbreak in China: using exponential attractor method in a spatial heterogeneous SEIQR model. Mathematical Biosciences and Engineering, 2020, 17(4): 3062-3087.
doi: 10.3934/mbe.2020174
[9]
Sarita Bugalia, Vijay Pal Bajiya, Jai Prakash Tripathi, Ming-Tao Li, Gui-Quan Sun .
Mathematical modeling of COVID-19 transmission: the roles of intervention strategies and lockdown. Mathematical Biosciences and Engineering, 2020, 17(5): 5961-5986.
doi: 10.3934/mbe.2020318
[10]
Qinghua Liu, Siyu Yuan, Xinsheng Wang .
A SEIARQ model combine with Logistic to predict COVID-19 within small-world networks. Mathematical Biosciences and Engineering, 2023, 20(2): 4006-4017.
doi: 10.3934/mbe.2023187
Abstract
The challenge for investors is the uncertainty of investing in a diversified portfolio using traditional techniques. This research project aimed to investigate and compare the effectiveness of two traditional Markowitz models and two machine learning techniques in optimizing diversified stock portfolios. The Markowitz models employed were the modern portfolio theory efficient frontier and quadratic mean-variance programming. The machine learning models utilized were the random forest classifier and the Extreme Gradient Boosting classifier. South African historical stock market time series listed companies on the Johannesburg Stock Exchange were analyzed to identify patterns and allocate stock weights to inform optimal stock portfolio diversification decisions. This minimizes risk while maximizing returns. Furthermore, the results are benchmarked against the JSE Top 40 Index. The portfolio generated by the eXtreme Gradient Boosting model performed the best, achieving a Sharpe ratio of 1.44. It is followed by the portfolio generated by the Random Forest classifier with a Sharpe ratio equal to 1.30. The portfolios generated by Modern Portfolio Theory Efficient Frontier and Quadratic Mean-Variance Programming had Sharpe ratios of 0.89 and 0.41, respectively. However, the JSE Top 40 Index portfolio had the lowest Sharpe ratio of 0.35. These results confirm the superiority of machine learning models over traditional Markowitz models. However, traditional models remain fundamental in understanding the principles of portfolio diversification. Machine learning models may be prioritized by investors primarily because of their ability to capture stochastic insights and market dynamics.
1.
Introduction
Mathematical models for the transmission dynamics of infectious diseases have been a powerful tool to understand and control epidemics. An important amount of these mathematical models are given by systems of ordinary differential equations (ODE's), considering a continuous-time framework. In the last decades, a great number of compartmental models has been proposed and many of them share some of the assumptions of the well known SIR model, first proposed by Kermack-McKendrick in 1927 [5] and also given by a system of ODE's.
The importance of compartmental models given by systems of ODE's has been even more highlighted since the beginning of the COVID-19 pandemic. In fact, SIR, SEIR, SEIRD-type models, among many others, have been used to analyze, predict and control the spread of SARS-CoV-2 virus worldwide, in what follows we refer to some of this models applied to COVID-19 given by systems of ODE's. The effect of the lockdown on the spread of COVID-19 was analyzed in [4] by considering a mathematical model that assesses the imposition of the lockdown in Nigeria. Besides lockdown measures, in [6] quarantine, and hospitalization of COVID-19 infected individuals is analyzed. A SIDARTHE compartmental model, is proposed in [10] which discriminates between infected individuals depending on whether they have been diagnosed and on the severity of their symptoms and is fitted to the COVID-19 epidemic in Italy. The authors claim that restrictive social-distancing measures will need to be combined with widespread testing and contact tracing to end the ongoing COVID-19 pandemic. The spread of COVID-19 in Portugal is modeled in [14], fitting the model to real data of simultaneously the number of active infected and hospitalized individuals, SEIR types models applied to COVID-19 are proposed in, for example, [12,16,21], were the models are fitted to real data. Quarantine and lockdown measures are considered in [24] with SEIQR type models. Fractional SEIR type models are also important in modeling and predicting the spread of infectious diseases. Their advantages in comparison to other type of models are highlighted in, for example, [1] were COVID-19 epidemic in Pakistan is analyzed through a fractional SEIR type model using the operator of Atangana-Baleanu. The Atangana-Baleanu derivative is also considered in [15]. A nabla discrete ABC-fractional order COVID-19 model is analyzed in [13]. In [2] a fractional model is proposed to study the first COVID-19 outbreak in Wuhan, China. The outbreak in Wuhan is also considered in [7,19]. A Caputo fractional model is proposed in [18] and fitted to COVID-19 data of Galicia, Spain and Portugal. A Bats-Hosts-Reservoir-People transmission fractional-order COVID-19 model for simulating the potential transmission with the thought of individual response and control measures by the government is analyzed in [23]. Stochastic epidemic models for COVID-19 spread and control are proposed in, e.g., [8,11]. A new class of distributions are applied to the generalized log-exponential transformation of Gumbel Type-II and implemented using real data of COVID-19, in [27]. The exponential transformation of Gumbel Type-II distribution for modeling COVID-19 data is used to analyze the number of deaths due to COVID-19 for Europe and China, in [28].
Among other important issues, in the mathematical analysis of compartmental models given by systems of ODE's, we may emphasize the stability analysis of the equilibrium points, the basic reproduction number and the model fitting to real data. Although the difficulty of this analysis depends on the complexity of the model under study, part of it is common to the majority of the models and may become simplified if we use adequate mathematical software. In this paper, we show how to use free and open-source software for the mathematical analysis of the models. We focus on the mathematical software SageMath (version 9.2) [22] and programming language [20], version 3.8, and the Python Libraries: Numpy (version 1.18.5), Pandas (version 0.25.3), Scipy (version 1.4.1) and Matplotlib (version 2.0.0).
In this note, we consider a SAIRP model, given by a system of five ODE's, for the transmission of SARS-CoV-2, first proposed in [26] and after generalized to piecewise constant parameters and complex networks model in [25]. We provide all the codes that allow us to compute the equilibrium points, basic reproduction number, visualize the global stability of the equilibrium points and by considering piecewise constant parameters. Moreover, we provide the Python code that allows to estimate some of the piecewise constant parameters and fit the model to the real data of COVID-19 transmission in Portugal, from March 2, 2020 until April 15, 2021. It is important to note, that all the codes are elementary and thus can be easily adapted to other compartmental models, see e.g., [3,6,9,17,24,30].
The paper is organized as follows. In Section 2, the SAIRP model, given by a system of ODE's, is described. The equilibrium points of the SAIRP model are computed in Section 3 and their global stability is illustrated using SageMath (version 9.2) [22] and Python (version 3.8) [20]. The generalized SAIRP model with piecewise constant parameters is presented and the Python code for the model fitting to COVID-19 Portuguese real data is given, in Section 4.
2.
Mathematical model
In this section, we consider the compartmental SAIRP mathematical model proposed in [25,26] for the transmission dynamics of COVID-19. We start by recalling the assumptions of the model. The total population , with (in days) and , is subdivided into five classes: susceptible individuals (); asymptomatic infected individuals (); active infected individuals (); removed (including recovered individuals and COVID-19 induced deaths) (); and protected individuals (). Therefore, , considering a continuous time framework, with . The total population is homogeneous and has a variable size, with constant recruitment rate, , and natural death rate, . The susceptible individuals become infected by contact with active infected and asymptomatic infected individuals, at a rate of infection , where represents a modification parameter for the infectiousness of the asymptomatic infected individuals and represents the transmission rate. Only a fraction of asymptomatic infected individuals develop symptoms and are detected, at a rate . Active infected individuals are transferred to the recovered/removed individuals , at a rate , by recovery from the disease or by COVID-19 induced death. A fraction , with , is protected (without permanent immunity) from infection, and is transferred to the class of protected individuals , at a rate . A fraction of protected individuals returns to the susceptible class , at a rate . Let and . The previous assumptions are described by the following system of ordinary differential equations:
(2.1)
The compartments and parameters descriptions and notations of system (2.1) are resumed in Table 1.
Table 1.
Description and notation of the compartments and parameters of model (2.1).
The model (2.1) is biologically and mathematically well-posed, that is for any initial condition , the system (2.1) admits a unique solution defined on , whose components are non-negative. Furthermore, the region defined by (2.2) is positively invariant [25].
3.
Equilibrium points and stability analysis
In this section, we use the free and open-source mathematical software SageMath (version 9.2) [22] to help us
compute the equilibrium points and the basic reproduction number of model (2.1). The global stability analysis of the equilibrium points is illustrated through numerical simulations developed using Python (version 3.8) [20].
3.1. Computing equilibrium points and basic reproduction number in SageMath
The model (2.1) has two equilibrium points:
● disease-free equilibrium, denoted by , given by
(3.1)
● endemic equilibrium, , whenever , given by
with
(3.2)
where the basic reproduction number, , is given by
Theorem 1 (Local stability of the DFE, [25]). The disease-free equilibrium, , is locally asymptoticallystable whenever .
Theorem 2 (Global stability of the DFE, [25]). If , then the disease-free equilibrium, , is globally asymptotically stable in .
Theorem 3 (Global stability of the EE, [25]). The compact region defined by
is positively invariant under the flow induced by system .It contains the disease-free equilibrium, , and the endemic equilibrium, , if .Furthermore, if , then the endemic equilibrium is globally asymptotically stable in .
The Python code to generate Figure 1 is given below. We first import several Python scientific libraries: contains standard routines for numerical computations; contains the function , which implements the Runge-Kutta method for integrating ODE's systems; contains useful functions for producing figures.
Next, we define the parameters for the SAIRP model (2.1).
Using expression (3.3), we easily compute the basic reproduction number .
Afterwards, we define the SAIRP model given by system (2.1).
Finally, we integrate the SAIRP model (2.1) with several randomly chosen initial conditions and we produce the 3D Figure 1.
4.
Mathematical model with piecewise constant parameters
In this section, we consider the SAIRP model with piecewise constant parameters, proposed in [25], which allows to model the impact of public health policies and the human behavior in the dynamics of the COVID-19 epidemic.
For the sake of simplicity, the equations of the SAIRP model (2.1) can be rewritten as
(4.1)
with and , where the non-linear operator is defined in by
(4.2)
The time line is subdivided into a finite number of intervals
with disjoint unions, and we introduce a piecewise constant function defined on each time interval as
with , and . Next, consider the sequence of Cauchy problems defined for each initial condition by
(4.3)
The following theorem establishes the well-posedness of system (4.3).
Theorem 4.For any initial condition , the sequence of Cauchyproblems given by model (4.3) admits a unique globalsolution, denoted again by , whose components are non-negative.Furthermore, the region is positively invariant [25].
We recall that the solutions of problem (4.3) are continuous on the time interval , but may not be of class at , . From the modeling point of view, each change of parameters occurring at time () corresponds, for example, to a public announcement of confinement/lift of confinement or prohibition of displacement [25].
In what follows, we sub-divide the time interval days into 9 sub-intervals and consider a set of piecewise parameters which are estimated in order to fit the real data of COVID-19 spread in Portugal, since the first confirmed case on March 2, 2020, until April 15, 2021.
Model fitting to real data with Python In order to fit the real data of active infected individuals by SARS-CoV-2 (detected by test) daily provided by the health authorities in Portugal [31], we use the model (4.3) with piecewise constant parameters and use the programming language Python (version 3.8) [20], and the Python Libraries: Numpy (version 1.18.5), Pandas (version 0.25.3), Scipy (version 1.4.1) and Matplotlib (version 2.0.0). The real data are available in [31] or, for example, in the data repository site, see e.g. [32]. In what follows we explain the goal of each code block.
First we need as previously to import several scientific libraries.
We consider the real data from [31,32], here denominated data-pt-covid19 with extension .xlsx. To read the real data one can use the following code.
Next, we subdivide the time window of the total 410 days into 9 subintervals. This specific subdivision is related to the increase/decrease of the transmission of the virus in the community, the public health measures implemented by the Portuguese authorities and the human behavior, that changes over time.
Next, we draw the curve of active infected individuals with COVID-19 in Portugal (real data), from the first confirmed case day, March 2, 2020, until April 15, 2021.
In order to solve the model (4.1) we need to define the initial conditions, for , that correspond to the number of individuals in each class on March 2, 2020.
Then, we define the values of the parameters that take constant values for all time .
The parameters , and are assumed to be piecewise constant. Moreover, we estimate the parameter for all days and the parameters and in some of the sub-intervals, using from Python, see [33] for more details, that uses non-linear least squares method to fit a function to data.
Finally, we make the plot with the real data and the model solution , for .
The graphic that results from the previous code is given in Figure 2.
Figure 2.
Output of the Python code: fitting the SAIRP model (4.3) to COVID-19 active infected individuals in Portugal, between March 02, 2020 and April 15, 2021.
We recall, that the mathematical method and numerical code can be applied to other models given by systems of ODE's that describe the transmission dynamics of different virus or bacteria in human, animals or cells, for example.
5.
Conclusions
In this note, we provided a complete toolkit for performing both a symbolic and numerical analysis of a recent epidemic model, presented in [25], introduced in order to study the spreading of the COVID-19 pandemic.
This innovative compartmental model takes into account the possible transmission of the virus by asymptomatic individuals, as well as the possibility to protect a fraction of the affected population by public health strategies, such as confinement or quarantine. Since the public health policies have a strong influence on the spreading of the epidemic, we have also considered a piecewise constant parameters extension of the initial autonomous system, which have proved its ability to fit with real data. Once the mathematical analysis of such a complex compartmental model can be tedious, we have presented a computational approach, which is intended to support the theoretical analysis:
● using the free and open-source software Sagemath, we have first computed the symbolic expressions of the basic reproduction number (Eq (3.3)), of the disease-free equilibrium (Eq (3.1)) and of the endemic equilibrium (Eq (3.2)), which highlight the role of each parameter of the model;
● using the scientific libraries numpy, and of the free and open-source language Python, we have integrated the epidemic model, in order to illustrate the theoretical stability statements (see Figure 1);
● finally, we fitted the piecewise constant parameters extension of the initial model with recent real-world data (see Figure 2).
Overall, the programs presented in this note have been written in a sufficiently general manner, so that they can easily be adapted to a great number of other epidemic models.
In a near future, we aim to apply our computational approach to an improved version of our complex compartmental model, so as to consider the effects of the mutations of the virus and the benefits of vaccination.
Acknowledgments
This research is partially supported by the Portuguese Foundation for Science and Technology (FCT) by the project UIDB/04106/2020 (CIDMA). Cristiana J. Silva is also supported by FCT via the FCT Researcher Program CEEC Individual 2018 with reference CEECIND/00564/2018.
Conflict of interest
The authors declare that they have no conflict of interest.
References
[1]
Abdi F, Abolmakarem S, Yazd AK, et al. (2024) Prospective portfolio optimization with asset preselection using a combination of long and short term memory and Sharpe ratio maximization. IEEE Access. https://doi.org/10.1109/ACCESS.2024.3466829
[2]
Arrow KJ, Debreu G (1954) Existence of an equilibrium for a competitive economy. Economet J Economet Soc, 265–290.
[3]
Attia EF, Aly SM, ElRawas As, et al. (2023) Portfolio diversification benefits before and during the times of covid-19: evidence from usa. Future Bus J 9: 26. https://doi.org/10.1186/s43093-023-00205-4 doi: 10.1186/s43093-023-00205-4
[4]
Avella A (2024) Real-world applications of Markowitz's portfolio optimization: A quantitative study. ResearchGate.
[5]
Breiman L (2001) Random forests. Mach learn 45: 5–32.
[6]
Chen T, Guestrin C (2016) Xgboost: A scalable tree boosting system. In Proceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining, 785–794. https://doi.org/10.1145/2939672.2939785
[7]
Deng GF, Lin WT, Lo CC (2012) Markowitz-based portfolio selection with cardinality constraints using improved particle swarm optimization. Expert Syst Appl 39: 4558–4566. https://doi.org/10.1016/j.eswa.2011.09.129 doi: 10.1016/j.eswa.2011.09.129
[8]
Fabozzi FJ, Kolm PN, Pachamanova DA, et al. (2007) Robust portfolio optimization and management. John Wiley & Sons.
[9]
Hastie T, Tibshirani R, Friedman JH, et al. (2009) The elements of statistical learning: data mining, inference, and prediction, 2. Springer. https://doi.org/10.1007/978-0-387-21606-5
[10]
Kumar RR, Ghanbari H, Stauvermann PJ (2024) Application of a robust maximum diversified portfolio to a small economy's stock market: An application to Fiji's south pacific stock exchange. J Risk Financ Manag 17: 388. https://doi.org/10.3390/jrfm17090388 doi: 10.3390/jrfm17090388
[11]
Markowitz H (1952) Portfolio selection. J Financ 7: 77–91.
[12]
Modigliani F, Miller MH (1958) The cost of capital, corporation finance and the theory of investment. Am Econ Rev 48: 261–297.
[13]
Murphy KP (2022) Probabilistic machine learning: an introduction. MIT press.
[14]
Nagurney A (2009) Portfolio optimization. Advanced Management Development Program in Real Estate.
[15]
Packard T, Gentilini U, Grosh M, et al. (2019) Protecting all: Risk sharing for a diverse and diversifying world of work. World Bank Publications.
[16]
Pandi A (2020) Mean-semivariance approach for portfolio optimisation.
[17]
Rathi V, Kshirsagar M, Ryan C (2024) Enhancing portfolio performance: A random forest approach to volatility prediction and optimization. In ICAART, 1278–1285.
[18]
García-Medina A, Rodríguez-Camejo B (2024) Random matrix theory and nested clustered optimization on high-dimensional portfolios. Int J Mod Phys C, 35: 1–19. https://doi.org/10.1142/S0129183124500980 doi: 10.1142/S0129183124500980
[19]
Sdg U (2019) Sustainable development goals. Energy Progress Report, Tracking SDG, 7: 805–814.
[20]
Sharpe WF (1966) Mutual fund performance. J Bus 39: 119–138.
[21]
Siew LW, Jaaman SH, Hoe LW (2019) Mathematical modelling of risk in portfolio optimization with mean-gini approach. In Journal of Physics: Conference Series, 1212: 012031. IOP Publishing.
[22]
Stiglingh ZC, Seitshiro MB (2022) Quantification of garch (1, 1) model misspecification with three known assumed error term distributions. J Financ Risk Manag 11: 549–578. https://doi.org/10.4236/jfrm.2022.113026 doi: 10.4236/jfrm.2022.113026
[23]
Sutiene K, Schwendner P, Sipos C, et al. (2024) Enhancing portfolio management using artificial intelligence: literature review. Front Artif Intell 7: 1371502. https://doi.org/10.3389/frai.2024.1371502 doi: 10.3389/frai.2024.1371502
[24]
Tan JHJ, Kek SL (2020) A simulation optimization model for portfolio selection problem with quadratic programming technique. In AIP Conference Proceedings, 2266. AIP Publishing.
[25]
Uykun FN (2024) Machine learning applications in portfolio optimization. Master's thesis, Middle East Technical University.
[26]
Van Greunen J, Heymans A (2023) Determining the impact of different forms of stationarity on financial time series analysis. In Business Research: An Illustrative Guide to Practical Methodological Applications in Selected Case Studies, 61–76. Springer.
[27]
Zanjirdar M (2020) Overview of portfolio optimization models. Adv Math financ Appl 5: 419–435. https://doi.org/10.22034/amfa.2020.1897346.1407 doi: 10.22034/amfa.2020.1897346.1407
[28]
Zhang C, Sjarif NNA, Ibrahim R (2024) Deep learning models for price forecasting of financial time series: A review of recent advancements: 2020–2022. Wiley Interdisciplinary Reviews: Data Mining and Knowledge Discovery, 14: e1519.
This article has been cited by:
1.
Azhar Iqbal Kashif Butt, Muhammad Rafiq, Waheed Ahmad, Naeed Ahmad,
Implementation of computationally efficient numerical approach to analyze a Covid-19 pandemic model,
2023,
69,
11100168,
341,
10.1016/j.aej.2023.01.052
2.
R. Prem Kumar, P.K. Santra, G.S. Mahapatra,
Global stability and analysing the sensitivity of parameters of a multiple-susceptible population model of SARS-CoV-2 emphasising vaccination drive,
2023,
203,
03784754,
741,
10.1016/j.matcom.2022.07.012
3.
Harald Øverby, Jan A. Audestad, Gabriel Andy Szalkowski,
Compartmental market models in the digital economy—extension of the Bass model to complex economic systems,
2023,
47,
03085961,
102441,
10.1016/j.telpol.2022.102441
4.
Manoj Kumar Singh, Brajesh K. Singh, Carlo Cattani,
Impact of general incidence function on three-strain SEIAR model,
2023,
20,
1551-0018,
19710,
10.3934/mbe.2023873
5.
Zenebe Shiferaw Kifle, Legesse Lemecha Obsu,
Mathematical modeling and analysis of COVID-19 and TB co-dynamics,
2023,
9,
24058440,
e18726,
10.1016/j.heliyon.2023.e18726
6.
Sofia Tedim, Vera Afreixo, Miguel Felgueiras, Rui Pedro Leitão, Sofia J. Pinheiro, Cristiana J. Silva,
Evaluating COVID-19 in Portugal: Bootstrap confidence interval,
2023,
9,
2473-6988,
2756,
10.3934/math.2024136
7.
Azhar Iqbal Kashif Butt, Waheed Ahmad, Muhammad Rafiq, Naeed Ahmad, Muhammad Imran,
Computationally efficient optimal control analysis for the mathematical model of Coronavirus pandemic,
2023,
234,
09574174,
121094,
10.1016/j.eswa.2023.121094
Esau Moyoweshumba, Modisane Seitshiro. Leveraging Markowitz, random forest, and XGBoost for optimal diversification of South African stock portfolios[J]. Data Science in Finance and Economics, 2025, 5(2): 205-233. doi: 10.3934/DSFE.2025010
Esau Moyoweshumba, Modisane Seitshiro. Leveraging Markowitz, random forest, and XGBoost for optimal diversification of South African stock portfolios[J]. Data Science in Finance and Economics, 2025, 5(2): 205-233. doi: 10.3934/DSFE.2025010