This first post attempts to point out some recent academic works on artificial intelligence (AI: defined as the ability of computers and machine to perform tasks that typically require human intervention) algorithms in finance mainly its quantitative side application to asset management such as: portfolio optimization, risk management, and forecasting. Academic works in other subsectors of finance such as: algorithmic trading; banking;insurrance; real estate; fintech, are beyond the scope of this short review.

Machine learning(a subset of AI) is a collection of algorithms that can have useful applications in many areas beside finance. Numerous reports, articles, and some books from elementary introductions to specialized deep analysis on applications and benefits of AI , ML and big data, regulatory and ethical issues, have been published over recent decades. Its specific use in finance has received a renewed attention from academia and practioners(individual investors, brokers,exchanges, fintech,etc) due to mainly increased computing power, availability of data, and recent advances in the quality of statistical models to deliver higher rate of forecasting accuracy. Most of these AI,ML applications try to implement, or improve on works produced years ago.

These supervised, unsupervised and deep learning(employs multiple layers to extract features from raw data) algorithms that have extensively studied in the literature include classification, regression, clustering, dimension reduction,just to name a few. Numerous ML packages, for instance, in R or Python programming language are readily available online.

The premise of combining AI and finance is to explore vast remaining opportunities that other traditional analysis tools may miss. Others hope that AI will help spread broad access to AI tools to build wealth.Therefore, some hail AI potential in financial field as the next transformative wave.

In the elite investment world such as hedge funds; investment firms have long adopted machine learning technology; and still relentlessly looking for new ways to gain an edge on one another. Obviously, AI finance applications can be highly specialized and complex; and its deplyoment or performance of their AI tools in real-world are not always made public. Also recently, many so called AI start-ups in wealth management; trading platforms, have sprung up. That said, no oracles just yet: these trading bots are, for now, still a ways from replicating human logic, dexterity or expertise.

The original idea was to hilight the crucial support of those diverse academic groups working behind the scene to advance AI use in financial field: mathematicians, statistitians, computer scientists, data scientists, engineers.

There’s a substantial literature out there on motivations of using and buliding robust AI techniques in general; and the benefits and chalenges of its use to model the complex, high-dimentional, volatile, noisy and nonlinear nature of financial data in particular. Recognizing that markets are not just quantitative in nature, but also subject to irrational and sychological behaviour, some add extra features such as sentiment analysis to enhance performance. Do the much-hyped AI algorithms really outperform so called traditional statistical models? At the core of it all is the quality of the statistical models.

In fact, some works investigate in-depth specific models i.e. running the same model under different assumptions; others compare or combine models or specific market micro-structure. Others ponder the question if AI algorithms can still handle black swan events or whether most popular or influential academic works can be replicated in real world i.e. if academics and practioners reach the same conclusions. As mentioned above, the literature is vast and varied. The following papers(and references within) among numerous others discuss various corners of machine learning in stock selection, portfolio allcation, prediction and trend in financial field; and much more. Some only provide literature review.Some provide code in R or Python.

So i started by asking some popular AI platforms what they have on the specific topic of AI, machine learning use in finance.

The folowing 4 papers is what i got when I asked Bing if it could generate a list or summary of academic research studies on the performance of machine learning in investment, portfolio allocation or risk management; and examples of institutions using machine learning in making investment decisions.

Bing answer: Based on the web search results, here are some of the studies that I found:

1. A review of machine learning experiments in equity investment decision-making: why most published research findings do not live up to their promise in real life:

This study critically evaluates 27 academic experiments that use machine learning to forecast equity markets and construct portfolios. It discusses the limitations and challenges of applying machine learning in real-world investment management, and provides recommendations for future research.

2. An analysis of machine learning risk factors and risk parity portfolio optimization:

This study examines the use of machine learning methods and elements to streamline a portfolio and allocate risk among different factors. It compares the performance of different machine learning algorithms and portfolio optimization techniques, and finds that the least variance portfolios outperform simpler benchmarks in minimizing risk.

3. Machine Learning Portfolio Allocation:

This study proposes a portfolio allocation strategy that employs machine learning to reward-risk time the market. It introduces a novel theoretical framework to explain the results, and shows that the strategy significantly improves investor utility and Sharpe ratios, and earns a large alpha of 4%.

4.Network and Machine Learning Approach to Factor, Asset, and Blended Allocation:

This study uses network analysis and machine learning to find asset and factor allocation solutions. It argues that factor exposure is desired within asset allocation, and demonstrates how machine learning can help investors achieve optimal diversification and risk-adjusted returns.

According to the web search results, some of these institutions are:

Hedge funds and broker-dealers: These are firms that use machine learning to find signals for higher and uncorrelated returns and optimize trading execution. Some examples of hedge funds that use machine learning are Renaissance Technologies, Two Sigma, and Bridgewater Associates.

Financial institutions and payments services: These are entities that use machine learning to spot fraudulent activity, assess credit quality, price insurance policies, and advance financial inclusion. Some examples of financial institutions and payments services that use machine learning are Adyen, Payoneer, Paypal, Stripe, and Skrill.

Public and private sector institutions: These are organizations that use machine learning for regulatory compliance, surveillance, data quality assessment, and fraud detection. Some examples of public and private sector institutions that use machine learning are the Federal Reserve Board, the Bank of England, the European Central Bank, and the International Monetary Fund.

A similar question, for example, in one of the papers:

Could present-day AI be a profitable investor in real world:

Bing answer:

Present-day AI can be a profitable investor in real-world investment, but it is not a guarantee. AI can help in investment management by augmenting the intelligence of human workers, developing new technologies, and analyzing large amounts of data. However, AI also faces some challenges and limitations, such as ethical issues, regulatory hurdles, and human biases.

Some examples of how AI can be a profitable investor are:

AI can use algorithmic trading to minimize transaction costs, improve order execution, and reduce human errors involved in trading securities.

AI can use machine learning to create models that can predict market trends, identify patterns, and optimize portfolios based on risk and return preferences.

AI can use deep learning to generate synthetic data and simulate scenarios that can help investors test their strategies and evaluate their performance.

Some examples of the challenges and limitations that AI faces are:

AI can pose ethical dilemmas, such as who is responsible for the outcomes of AI decisions, how to ensure fairness and transparency in AI algorithms, and how to protect the privacy and security of data used by AI.

AI can encounter regulatory barriers, such as how to comply with existing laws and regulations that may not be designed for AI applications, how to deal with cross-border issues and jurisdictional conflicts, and how to adapt to changing rules and standards.

AI can be influenced by human biases, such as how to avoid replicating or amplifying the prejudices and errors of human data sources, how to account for the uncertainty and complexity of real-world situations, and how to balance the trade-off between efficiency and explainability.

Therefore, present-day AI can be a profitable investor in real-world investment, but it also requires careful design, implementation, and oversight. AI is not a magic bullet that can guarantee success, but rather a powerful tool that can enhance human capabilities and creativity.

Same question to ChatGPT: Could present-day AI be a profitable investor in real world:

ChatGPT answer:

Present-day AI, including language models like myself, is not designed for direct financial investment or trading. While AI and machine learning algorithms have made significant strides in various areas of finance, such as algorithmic trading and portfolio optimization, there are several important considerations to keep in mind:

Data Quality: AI models heavily rely on data. If the data used for training and decision-making is flawed or incomplete, it can lead to inaccurate predictions and financial losses.

1.Model Risk: AI models are not infallible. They can make errors, especially during unexpected market events or when operating outside the scope of their training data.

2.Market Dynamics: Financial markets are influenced by a multitude of factors, including geopolitical events, economic indicators, and human psychology. AI models may struggle to adapt to rapidly changing conditions.

3.Regulatory Compliance: Financial markets are heavily regulated, and trading algorithms must comply with strict rules. Developing AI-driven trading systems that adhere to these regulations is a complex task.

4.Overfitting: There’s a risk of overfitting AI models to historical data, making them perform well in the past but poorly in the future when market conditions change.

5.Ethical Considerations: AI models should also consider ethical concerns, such as avoiding discriminatory lending practices or socially irresponsible investments.

While AI can certainly assist human investors by providing data analysis, pattern recognition, and risk assessment, it’s not a guaranteed way to achieve consistent profitability in financial markets. Successful investment and trading require a combination of financial expertise, risk management, and an understanding of the broader economic and geopolitical landscape. Additionally, human judgment and decision-making are still essential components in navigating the complexities of financial market.

Same question to Bard: could present-day AI be a profitable investor in real world?

Bard answer:

Yes, present-day AI can be a profitable investor in the real world. AI has been shown to be able to outperform human investors in some cases, and it is still under development. As AI technology continues to improve, it is likely that AI investors will become even more profitable.

Here are some of the ways that AI can be used for investing:

2. Machine learning: AI can use machine learning to learn from past data and make predictions about future market movements. This can help investors to make better investment decisions.

3. Algorithmic trading: AI can be used to develop algorithmic trading strategies that can automatically buy and sell stocks based on pre-programmed rules. This can help investors to take advantage of market opportunities quickly and efficiently.

Of course, there are also some challenges to using AI for investing. One challenge is that AI models can be biased, which can lead to bad investment decisions. Another challenge is that AI models can be complex and difficult to understand, which can make it difficult for investors to trust them.

Overall, the potential for AI to be a profitable investor is very promising. As AI technology continues to improve, it is likely that AI investors will become even more profitable. However, it is important to be aware of the challenges of using AI for investing and to take steps to mitigate those challenges.

Here are some additional things to consider when investing with AI:

*.The type of AI model used. Some models are better suited for certain types of investments than others.

*. The amount of data used to train the model. More data generally leads to better results, but it can also make the model more complex and difficult to understand.

* The risk tolerance of the investor. Some AI models are more aggressive than others, and it is important to choose a model that is appropriate for the investor’s risk tolerance.

* The cost of using the AI model. Some models are free to use, while others require a subscription fee.

If you are considering using AI for investing, it is important to do your research and choose a reputable provider. You should also be aware of the risks involved and take steps to mitigate those risks.

Question :can you suggest some academic studies where AI algorithms have outperformed some elite investors?

Bard answer:

Sure, here are some academic studies where AI algorithms have outperformed some elite investors:

1.”Can Machine Learning Outperform Human Professional Investors? Evidence from the S&P 500” by Jun Pan, Lingfeng Yang, and Ming Ding (2020). This study found that an AI algorithm that used machine learning to predict stock prices outperformed the S&P 500 index by an average of 1.5% per year over a 10-year period.

2.”The Rise of Artificial Intelligence in Financial Trading” by Joseph H. Wang, Wei-Ling Hsu, and Yi-Hsuan Yang (2020). This study found that AI algorithms are increasingly being used in financial trading, and that they have the potential to outperform human traders.

3.”Artificial Intelligence in Asset Management: A Survey” by Yuval Heller, Amir Lichtenstein, and Ronen Feldman (2020). This study provides a comprehensive overview of the use of AI in asset management, and discusses the challenges and opportunities that AI presents to the industry.

4.”Deep Reinforcement Learning for Financial Trading” by Mingyang Zhang,Li Zhang, and Lingxiao Zhou (2020). This study developed a deep reinforcement learning algorithm that can learn to trade stocks profitably. The algorithm outperformed a human-designed trading algorithm by an average of 5% per year over a 10-year period.

5.”A Systematic Review of Artificial Intelligence in Finance” by Yifan Zhang, Feifei Gao, and Yizhou Sun (2021). This study reviewed the academic literature on the use of AI in finance. The study found that AI has the potential to improve investment performance, risk management, and operational efficiency in the financial industry.

These are just a few examples of academic studies that have shown the potential of AI to outperform human investors. As AI technology continues to improve, it is likely that AI will play an increasingly important role in the financial industry.

Bard:

Sure, here are some more specific deep learning tools used in the elite investment world:

* Convolutional neural networks (CNNs): CNNs are a type of deep learning algorithm that are used for image recognition. CNNs are used in a variety of investment applications, such as analyzing financial charts and identifying patterns in market data.

* Long short-term memory (LSTM) networks: LSTMs are a type of RNN that are specifically designed to handle long sequences of data. LSTMs are used in a variety of investment applications, such as predicting stock prices and identifying trading opportunities.

* Generative adversarial networks (GANs): GANs are a type of deep learning algorithm that are used for generating new data. GANs are used in a variety of investment applications, such as creating synthetic financial data and testing trading strategies.

* Autoencoders: Autoencoders are a type of deep learning algorithm that are used for dimensionality reduction. Autoencoders are used in a variety of investment applications, such as extracting features from market data and clustering stocks.

These are just a few of the many deep learning tools that are used in the investment world. The specific tools that are used will vary depending on the specific investment application.

It is important to note that deep learning tools are not always a guarantee of success in the investment world. There are a number of factors that can affect the performance of a deep learning tool, such as the quality of the data used to train the tool, the complexity of the tool, and the market conditions.

Despite these challenges, deep learning is a powerful tool that can be used to improve investment performance. As deep learning technology continues to improve, it is likely that deep learning will play an increasingly important role in the investment world.

Bard may display inaccurate or offensive information that doesn’t represent Google’s views.

Academic papers:

Some machine learning academic papers on the performance of machine learning in investment portfolio allocation or risk management. Some contain R or Python code. source:ELICIT.ai

SUMMARY OF TOP 4 PAPERS

BETA

These papers provide a comprehensive review of the application of machine learning algorithms to investment, portfolio allocation, and risk management with code in R programming.

Wang (2023) discusses the use of reinforcement learning, recurrent neural networks, and random forests in portfolio optimization.

Asawa (2022) presents a review of several machine learning portfolio optimization techniques, including clustering-based, support vector machines-based, and genetic algorithm-based. Al-Muharraqi (2023) provides a literature review of implementing machine learning in optimizing stock portfolios, highlighting the most applied machine learning algorithm in predicting stock returns and the most adopted approach for portfolio optimization. Finally, Untwal (2020) applies machine learning for financial forecasting using supervised and unsupervised learning in the form of data frame with multiple regression model and factor analysis. Overall, these papers provide a comprehensive overview of the application of machine learning algorithms to investment, portfolio allocation, and risk management with code in R programming, highlighting the most effective techniques and algorithms for optimizing portfolios and forecasting financial outcomes.

Paper title

Abstract summary

1. Review: Application of Machine Learning to Investment Portfolios

2023

Machine learning methods to build an optimal portfolio can help investors minimize risk and maximize returns.

2.Modern Machine Learning Solutions for Portfolio Selection

Yash S Asawa

IEEE Engineering Management Review

2022

Machine learning models have seen tremendous success in the portfolio selection problem as well.

3. Implementing Machine Learning in Optimizing Stock Portfolios: A review

Maitham Al-Muharraqi, M. Messaadia

2023 International Conference On Cyber Management And Engineering (CyMaEn)

2023

Long short-term memory model was the most applied machine learning algorithm in predicting stocks returns.

4.Application of machine learning using R-programming for financial forecasting

Nitin Untwal

2020

Unsupervised learning is applied by way of factor analysis to group the variables used in the regression model.

5.A machine learning approach to portfolio pricing and risk management for high‐dimensional problems

Lucio Fernandez-Arjona, D. Filipović

2020

A replicating martingale is learned from a finite sample in a supervised setting.

6.Predicting the Link between Stock Prices and Indices with Machine Learning in R Programming Language

Mengya Cao

Journal of mathematics

2021

The proposed buying and selling strategies and stock price trend prediction methods in this paper are effective.

7.Evaluating the Performance of Machine Learning Algorithms in Financial Market Forecasting: A Comprehensive Survey

Lukas Ryll, Sebastian Seidens

2019

Recurrent neural networks outperform feed-forward neural networks and support vector machines in financial market forecasting.

8.Portfolio Optimization using Artificial Intelligence: A Systematic Literature Review

G. C. Santos, Flavio Barboza, Antônio C. P. Veiga, Kamyr Gomes

Exacta

2022

Artificial intelligence models can help investors find portfolios in which the focus is to optimize the risk-return relationship.

9.Some Observations for Portfolio Management Applications of Modern Machine Learning Methods

Jooyoung Park, Seongman Heo, Taehwan Kim, Jeongho Park, Jaein Kim, Kyungwook Park

International Journal of Fuzzy Logic and Intelligent Systems

2016

Modern data-driven machine learning methods for portfolio management problems are useful in practical application domains.

10.Analysis of some portfolio selection models for R&D

A. Gear, A. Lockett, A. Pearson

1971

11.Machine Learning Algorithms to Classify Future Returns Using Structured and Unstructured Data

J. Livnat, J. Singh

2021

Machine learning improves the accuracy of predictions in the three classes to a degree that is statistically very significant when judged against random classifications.

12.mlr: Machine Learning in R

B. Bischl, Michel Lang, Lars Kotthoff, J. Schiffner, Jakob Richter, Erich Studerus, Giuseppe Casalicchio, Zachary M. Jones

Journal of machine learning research

2016

The MLR package provides a generic, object-oriented, and extensible framework for classification, regression, survival analysis, and clustering for the R language.

13.Research on Quantitative Investment Based on Machine Learning

K. Zhang

2020

Machine learning methods to construct quantitative investment strategies are obvious advantages.

14.I. Machine Learning and Data Science Applications in Investments

Larry Cao

2023

AI and big data tools can augment your existing investment processes.

Optimization algorithms and investment portfolio analytics with machine learning techniques under time-varying liquidity constraints

Mazin A. M. Al Janabi

2021

The recommended reinforcement machine learning optimization algorithms can aid in solving some real-world dilemmas under stressed and adverse market conditions.

15.A Review on Machine Learning for Asset Management

Pedro M. Mirete-Ferrer, Alberto Garcia-Garcia, J. Baixauli-Soler, Maria A. Prats

Risks

2022

Existing methods are reviewed, highlighting their contribution and significance in the analyzed financial disciplines.

16.Portfolio optimization with return prediction using deep learning and machine learning

Yilin Ma, Ruizhu Han, Weizhong Wang

Expert Syst. Appl.

2021

The mean-variance model with random forest return prediction outperforms the other models.

17.A Machine Learning Integrated Portfolio Rebalance Framework with Risk-Aversion Adjustment

Zhenlong Jiang, Ran Ji, Kuo-Chu Chang

Journal of Risk and Financial Management

2020

The XGBoost model provides the best prediction of market movement.

18.A Comprehensive Review of Machine Learning for Financial Market Prediction Methods

Rahul Maruti Dhokane, O. P. Sharma

2023 International Conference on Emerging Smart Computing and Informatics (ESCI)

2023

Artificial neural network and support vector machine techniques are used for forecasting the financial market.

19.A Machine Learning Approach to Research Curation for Investment Process

Stephen Lawrence, Sonya Cates, Carla Penedo, V. Samatova

2016

Machine learning approach approximates the quality of human curation.

20.Machine Learning Stock Market Prediction Studies: Review and Research Directions

T. Strader, J. Rozycki, T. H. Root, Yu-Hsiang Huang

Journal of International Technology and Information Management

2020

Artificial neural network studies, support vector machine studies, and hybrid or other artificial intelligence approaches are categories of machine learning stock market prediction research.

21.Asset Allocation via Machine Learning and Applications to Equity Portfolio Management

Qing Yang, Zhenning Hong, Ruyan Tian, Tingting Ye, Liangliang Zhang

2020

A machine learning-based bottom-up approach for static and dynamic portfolio optimization on potentially a large number of assets overcomes many major difficulties arising in current optimization schemes.

22.Decision Support Using Machine Learning Indication for Financial Investment

Ariel Vieira de Oliveira, Márcia Cristina Schiavi Dazzi, Anita M. R. Fernandes, R. Dazzi, Paulo Ferreira, V. Leithardt

Future Internet

2022

The Decision Tree was the algorithm that obtained the best classification metrics and an accuracy of 77%.

23.Study of machine learning algorithms for potential stock trading strategy frameworks

Aakash Agarwal

International Journal of Financial Accounting and Management

2021

The random forest algorithm is the most suitable algorithm for forecasting a stock’s market prices based on various data points from historical data.

24.Machine Learning Stock Market Prediction Studies: Review and Research Directions

T. Strader, J. Rozycki, T. H. Root, Yu-Hsiang Huang

Journal of International Technology and Information Management

2020

Artificial neural network studies, support vector machine studies, and hybrid or other artificial intelligence approaches are categories of machine learning stock market prediction research.

25. Application of Machine Learning Algorithms in Stock Market Prediction

Sumit Kumar, Sanlap Acharya

2020

Long Short-Term Memory algorithm performed better than the others among the supervised learning techniques.

26.Literature review: Machine learning techniques applied to financial market prediction

Bruno Miranda Henrique, Vinicius Amorim Sobreiro, Herbert Kimura

Expert Syst. Appl.

2019

The research theme is still relevant.

27.Study of Machine learning Algorithms for Stock Market Prediction

Ashwin Pathak, Sakshi Pathak

2020

The successful prediction of the stock market will have a very positive impact on the stock market institutions and the stock market.

28.Modern Probabilistic Machine Learning and Control Methods for Portfolio Optimization

Jooyoung Park, Jungdong Lim, Wonbu Lee, Seung-Hyun Ji, Keehoon Sung, Kyungwook Park

International Journal of Fuzzy Logic and Intelligent Systems

2014

Probabilistic machine learning and control-based solutions to a couple of portfolio optimization problems work well when applied to real financial market data.

29.Machine Learning for Financial Engineering (Advances in Computer Science and Engineering: Texts)

L. Györfi, G. Ottucsák, Harro Walk

2011

Sequential investment strategies use information collected from the market’s past and determine a portfolio at the beginning of a trading period.

30.Model-based Deep Reinforcement Learning for Dynamic Portfolio Optimization

Pengqian Yu, J. Lee, Ilya Kulyatin, Zekun Shi, Sakyasingha Dasgupta

arXiv.org

2019

A model-based approach works with both on-policy or off-policy RL algorithms.

31.A Machine Learning Framework for Stock Selection

XingYu Fu, JinHong Du, Yifeng Guo, MingWen Liu, Tao Dong, XiuWen Duan

arXiv.org

2018

Stacking outperforms other models reaching an AUC score of 0.972.

32.Machine Learning Algorithms for Financial Asset Price Forecasting

Philip Ndikum

ArXiv

2020

Implemented machine learning models significantly outperform the CAPM on out-of-sample test data.

33.Review on Machine Learning Techniques For Stock Prediction

R. Shah, Dhruva Patodia

2018

Machine learning is mostly in use to demonstrate the promise of producing consistently accurate estimates without needing multiple manual edits to the program.

34.Machine learning applied to stock index performance enhancement

Tien-Yu Hsu

J. Bank. Financial Technol.

2021

The stock selection model can measure import features appropriately.

35.Study on Machine Learning Techniques In Financial Markets

Prakhar Vats, Krishna Samdani

2019 IEEE International Conference on System, Computation, Automation and Networking (ICSCAN)

2019

Different trading techniques are introduced and their effectiveness in quantitative trading and, in general, finance to generate alphas is observed.

36.Quantitative Investment with Machine Learning in US Equity Market

Yuxiang Huang

2019

The annual return is above 40%.

37.Risk-Aware Reinforcement Learning for Multi-Period Portfolio Selection

David Winkel, Matthias Schubert, T. Seidl

2022

A single meta agent generates optimized portfolio allocation policies for any level of risk preference in a given interval.

38.Stock Prediction using Machine Learning a Review Paper

Nirbhey Singh Pahwa, Neeha Khalfay, Vidhi Soni, D. Vora

2017

Review

A machine learning application focusing on the patterns generated in the BSE stock trade over the period of time is essential.

39.Application and Comparison of Multiple Machine Learning Models in Finance

Yali Jiang

Scientific Programming

2022

Stocks with higher scores have higher investment value.

40.A Fund Selection Robo-Advisor with Deep-learning Driven Market Prediction

Chen-Sheng Gu, H. Hsieh, Chung-Shu Wu, R. Chang, Jan-Ming Ho

IEEE International Conference on Systems, Man and Cybernetics

2019

The accuracy of our market prediction method can reach 84.3%.

41.Research on Quantitative Investment Strategies Based on Deep Learning Algorithms in the Context of the Need for Information Management

Yueheng Wang, Shaohang Huang

International Conference on Information Management

2022

The quantitative investment method proposed in this paper can achieve higher investment returns compared with other methods.

42.Introduction and Implementation of Machine Learning Algorithms in R

S. R. M. Sekhar, G. Siddesh

Research Anthology on Artificial Intelligence Applications in Security

2019

Machine learning algorithms are present in computer science.

43.Integration of Macroeconomic Data into Multi-Asset Allocation with Machine Learning Techniques

Amine Abouseir, Arthur Le Manach, Mohamed El Mennaoui, Ban Zheng

2020

Machine learning algorithms to our investment universe and then apply different portfolio allocation methods.

44.Application of Machine Learning Algorithms in the Stock Market Analysis

Chunjiang Li

Highlights in Business, Economics and Management

2023

Machine learning algorithms have helped investors make improvements in stock market analysis.

45.A Novel Anti-Risk Method for Portfolio Trading Using Deep Reinforcement Learning

Han Yue, Jiapeng Liu, Dongmei Tian, Q. Zhang

Electronics

2022

A novel anti-risk portfolio trading method based on deep reinforcement learning is effective and superior to the Dow Jones Industrial Average index.

46.Reinforcement learning in financial markets - a survey

Thomas G. Fischer

2018

The RL research community has made considerable advances in the finance domain over the past two decades.

47.Machine learning for stock selection

Robert J. Yan, C. Ling

Knowledge Discovery and Data Mining

2007

The Prototype Ranking method leads to a clear profit improvement.

48.Machine Learning Algorithms in Stock Market Prediction

Jayesh Potdar, Rejo Mathew

Lecture Notes on Data Engineering and Communications Technologies

2019

Support vector regression, improved Levenberg Marquardt algorithm, and linear regression are machine learning algorithms for stock prediction.

49.A risk-aware approach to stock portfolio allocation based on Deep Q-Networks

Jacopo Fior, Luca Cagliero

Advanced Industrial Conference on Telecommunications

2022

The proposed system performs better than state-of-the-art RL solutions in terms of cumulative return, stability, and sharpe ratio.

50.A Comparative Study of Machine Learning Algorithms for Financial Data Prediction

Bencharef Omar, G. Daniel, Bousbaa Zineb, Cortés Jofré Aida

International Symposium Advanced Electrical and Communication Technologies

2018

Machine learning has been widely used as a part of stock market investment strategies.

51.Dynamic portfolio rebalancing through reinforcement learning

Qing Yang Eddy Lim, Qi Cao, Hiok Chai Quek

Neural Comput. Appl.

2022

The proposed reinforcement learning agent for gradual portfolio rebalancing with the Long Short-Term Memory model on price prediction outperforms the other three methods.

52.A Practical Machine Learning Approach for Dynamic Stock Recommendation

Hongyang Yang, Xiao-Yang Liu, Qingwei Wu

2018 17th IEEE International Conference On Trust, Security And Privacy In Computing And Communications/ 12th IEEE International Conference On Big Data Science And Engineering (TrustCom/BigDataSE)

2018

The proposed scheme outperforms the long-only strategy on the S&P 500 index in terms of Sharpe ratio and cumulative returns.

53.Application of machine learning in stock selection

Pengfei Li, Jungang Xu, Mohammad AI-Hamami

Applied Mathematics and Nonlinear Sciences

2022

The stock selection model constructed by these six machine learning methods can obtain higher return and stability.

54.A Comparative Study of Supervised Machine Learning Algorithms for Stock Market Trend Prediction

I. Kumar, Kiran Dogra, Chetna Utreja, Premlata Yadav

International Conference Inventive Communication and Computational Technologies

2018

The Random Forest algorithm performs the best for large datasets.

55.Machine Learning in Asset Pricing

S. Nagel

2021

Machine learning methods are quickly becoming part of the toolkit in asset pricing research and quantitative investing.

56.Introduction and Implementation of Machine Learning Algorithms in R

S. R. Mani Sekhar, G. M. Siddesh

Research Anthology on Artificial Intelligence Applications in Security

2021

Machine learning algorithms are present in computer science.

57.Building portfolios based on machine learning predictions

Tomasz Kaczmarek, Katarzyna Perez

Economic Research-Ekonomska Istraživanja

2021

Mean-variance and hierarchical risk parity optimizers outperform the 1/N rule for portfolio weight creation using returns of stocks from the S&P500 and STOXX600 for robustness.

58.A portfolio recommendation system based on machine learning and big data analytics

Man-Fai Leung, Abdullah Jawaid, Sai-Wang Ip, Chun-Hei Kwok, Shing Yan

Data Science in Finance and Economics

2023

A machine learning-based portfolio recommendation system and big data analytics can effectively meet the expectations of the majority of users and enhance users’ financial knowledge.

59.Tradinnova-LCS: Dynamic stock portfolio decision-making assistance model with genetic based machine learning

Isidoro J. Casanova

IEEE Congress on Evolutionary Computation

2010

The best stocks are selected to be incorporated in the portfolio by reinforcement learning with genetic programming daily by the system.

60.Machine Learning for Stock Selection

K. Rasekhschaffe, Robert C. Jones

Financial analysts journal

2019

Machine learning algorithms can uncover subtle, contextual, and nonlinear relationships.

61.Machine learning algorithms for time series in financial markets

M. Ghasemzadeha, N. Mohammad-Karimi, habib ansari-samani

2020

The selected technical features by the leading methods could find the best and most efficient values for the parameters of the algorithms.

62.Machine Learning for Quantitative Finance Applications: A Survey

F. Rundo, F. Trenta, Agatino Luigi Di Stallo, S. Battiato

Applied Sciences

2019

Machine learning techniques in the field of quantitative finance outperform traditional approaches.

63.Portfolio construction based on stock returns prediction using machine learning

Sinan Liu, Zihan Wu, Xianyi Zhang

Other Conferences

2022

The expected return generated by LSTM was used to construct the portfolio model by allocating risks and returns.

64.Asset Price Prediction via Machine-Learning Method: A Review

Yajie Yang, Bingxu Chen, L. Zhang

International Conference on Computational Intelligence and Security

2021

The best-performing methods with nonlinear predictor interactions are random forests and neural networks.

65.Combining Reinforcement Learning and Inverse Reinforcement Learning for Asset Allocation Recommendations

I. Halperin, Jiayu Liu, Xiao Zhang

2022

The inverse reinforcement learning component learns the intent of fund managers as suggested by their trading history.

66.Supervised classification-based stock prediction and portfolio optimization

Sercan Ö. Arik, S. Eryilmaz, Adam Goldberg

ArXiv

2014

The portfolio our system suggests by predicting the behavior of stocks results in a 3% larger growth on average than the overall market within a 3-month time period.

67.Design and Research of Intelligent Quantitative Investment Model Based on PLR-IRF and DRNN Algorithm

Chunming Tang, X. Zheng, Xiang Yu, Chunkai Chen, Wenyan Zhu

2018 IEEE 4th Information Technology and Mechatronics Engineering Conference (ITOEC)

2018

The prediction accuracy of this model is relatively high.

68.Machine Learning in Finance

Dhruvil Trivedi, Ashish Bhagchandani, Rutul Ganatra, M. Mehta

2018 IEEE Punecon

2018

The proposed algorithm is to create an impact in the near future and to make easy for humans to study the financial market.

69. Overview of Machine Learning for Stock Selection Based on Multi-Factor Models

Haoxuan Li, Xueya Zhang, Ziyang Li, Chunyuan Zheng

E3S Web of Conferences

2020

Machine learning algorithms perform better on stock selection than traditional statistical methods in multi-factor models.

70.Machine Learning in Asset Management—Part 1: Portfolio Construction—Trading Strategies

Derek Snow

2019

Machine learning can help with most portfolio construction tasks like idea generation, alpha factor design, asset allocation, weight optimization, position sizing, and the testing of strategies.

71. Machine Learning in Asset Management—Part 2: Portfolio Construction—Weight Optimization

Derek Snow

2020

Machine learning can help with most portfolio construction tasks.

72.MA-FDRNN: Multi-Asset Fuzzy Deep Recurrent Neural Network Reinforcement Learning for Portfolio Management

Tarrin Skeepers, Terence L van Zyl, A. Paskaramoorthy

2021 8th International Conference on Soft Computing & Machine Intelligence (ISCMI)

2021

The MA-FDRNN method can take short positions.

73.Importance of Machine Learning in Making Investment Decision in Stock Market

Akhilesh Prasad, A. Seetharaman

Vikalpa The Journal for Decision Makers

2021

The performance of models measured by root-mean-square error for regression and accuracy score for classification models varied greatly.

74.Introduction to special issue on machine learning for business applications

C. Ling

TIST

2011

A dual-classification method for the contrarian trading strategy can form better portfolios.

75.A Novel Machine Learning Strategy Based on Two-Dimensional Numerical Models in Financial Engineering

Qingzhen Xu

2013

The proposed machine learning method uses partial differential equations to predict the stock market.

76.An Overview of Machine Learning, Deep Learning, and Reinforcement Learning-Based Techniques in Quantitative Finance: Recent Progress and Challenges

S. Sahu, A. Mokhade, N. Bokde

Applied Sciences

2023

Deep reinforcement learning algorithms are used to anticipate and evaluate stock and foreign exchange markets.

77.A Comparative Study on Machine Learning Techniques in Assessment of Financial Portfolios

Nikhitha Pai, V. Ilango

International Conference on Communication and Electronics Systems

2020

Machine learning methods are used in the area of stock selection for portfolio construction and portfolio management.

78.Reinforcement learning in Portfolio Management and its interpretation

L. Weijs

2018

Reinforcement learning is perfectly suited for the volatile dynamic programming environment of portfolio management compared to other machine learning methods in use.

79.Understanding Machine Learning for Diversified Portfolio Construction by Explainable AI

Markus Jaeger, Stephan Krügel, D. Marinelli, J. Papenbrock, P. Schwendner

2020

Features addressing the hierarchical properties of the universe are connected to the relative performance of HRP respect to ERC in the augmented dataset built from a multi-asset investment universe of commodities, equities and fixed income futures.

80.Model-based Deep Reinforcement Learning for Financial Portfolio Optimization

Pengqian Yu, J. Lee, Ilya Kulyatin, Zekun Shi, Sakyasingha Dasgupta

2019

A proposed model is robust, profitable and risk-sensitive.

81.RPS: Portfolio Asset Selection using Graph based Representation Learning

MohammadAmin Fazli, Parsa Alian, A. Owfi, Erfan Loghmani

arXiv.org

2021

A heuristic for getting closer to the optimal subset of assets can be used for portfolio selection to increase diversification.

82.Forecasting ETFs with Machine Learning Algorithms

J. Liew, Boris Mayster

2017

ETF returns can be predicted with machine learning algorithms.

83.Machine Learning in Empirical Asset Pricing Models

Huei-Wen Teng, Yu-Hsien Li, S. Chang

2020 International Conference on Pervasive Artificial Intelligence (ICPAI)

2020

Neural networks produce portfolios with the highest Sharpe ratios for both buy-and-hold and long-short strategies.

84.Sequential machine learning approaches for portfolio management

Nicolas Chapados

2009

A functional representation of time series allows forecasts to be performed over an unspecified horizon with progressively-revealed information sets.

85.Machine Learning in Asset Management

Derek Snow

SSRN Electronic Journal

2019

The informal paper will organically grow with future developments in machine learning and data processing techniques.

86.Reward Adjustment Reinforcement Learning for Risk-averse Asset Allocation

Jian Li, L. Chan

The 2006 IEEE International Joint Conference on Neural Network Proceedings

2006

Most classical reinforcement learning algorithms do not take risk into account.

87.The application of artificial intelligence investment in capital markets: A case study of two constituent stocks of Dow Jones

D. Jones, Wei-Yuan Lin, Yuan-De Chu

2020

Deep learning techniques can be further adopted in the future to obtain better accuracy and faster research results.

88.Machine Learning with R

Brett Lantz

Machine-mediated learning

2013

Machine learning methods to a variety of common tasks including classification, prediction, forecasting, market basket analysis, and clustering.

89.A Review on Applied Data Mining Techniques to Stock Market Prediction

N. Fidan, B. A. Ozkok

2013

Hybrid models that combine data mining techniques or integrate an algorithm to a method work efficiently.

90.Online machine learning algorithms for currency exchange prediction

Eleftherios Soulas, D. Shasha

2013

Software like this may enhance the trader’s performance.

91.Asset Allocation via Machine Learning

Zhenning Hong, Ruyan Tian, Qing Yang, Wei Yao, Tingting Ye, Liangliang Zhang

Accounting and Finance Research

2021

A machine learning-based numerical framework solves static and dynamic portfolio optimization problems with potentially an extremely large number of assets.

92.A Study of Stock Portfolio Strategy Based on Machine Learning

Zhuoyuan Ouyang