Paper/Article 1 A Comparative Bhattacharjee 2019 This paper's main This survey is Study between and P. objective is to compare useful in Traditional Bhattacharja, the effectiveness of understandin Statistical machine learning g the needs, Approach and techniques and fundamentals, Machine conventional statistical frameworks, Learning methodologies. In and Approach terms of prediction techniques for accuracy and stock market performance, a trend comparison of analysis. statistical and machine learning methodologies has been conducted. Machine learning methods are most effective in predicting stock prices for low MSE and MAPE values. 2 Stock Market Vanukuru, 2018 The prediction of a Prediction Using Kranthi stock using machine Machine learning is explained in Learning this study. The majority of stockbrokers employ technical, fundamental, or time series analysis when making stock predictions. This study employs prices with both daily and instantaneous frequencies and a machine learning technique called Support Vector Machine (SVM). 3 Stock Closing Vijh, Mehar 2020 Artificial Neural Price Prediction Network and Random using Machine Forest methodologies Learning are used for this study. Techniques The models are effective in forecasting stock closing price, as seen by the low values of RMSE and MAPE. 4 Stock Closing Vijh, Mehar 2019 This project's ultimate Price Prediction goal is to forecast using Machine Bombay Stock Learning Exchange behaviour. Techniques They have considered variables such as Commodity Prices (crude oil, gold, and silver), Market History, and Foreign Exchange Rate (FEX). Their results demonstrate that BSE is most dependent on the gold rate. 5 Stock Prediction A M Pranav 2022 This study examines using Machine the many methods Learning used to forecast share values, including classical machine learning, deep learning, neural networks, and graph-based methods. The study was expanded to include Deep Learning and Neural Network models such as Convolutional Neural Networks, Long Short Term Memory, etc. 6 Machine Soni, Payal 2021 Intriguing study has Learning been done on the topic Approaches in of stock market Stock Price forecasting using Prediction Twitter sentiment analysis. Found a strong link between stock market movements and public mood as reflected on Twitter. Their work's key contribution is the creation of a sentiment analyser that can determine what kind of sentiment is present in a tweet. 7 Sentiment K. N. Reddy 2018 To get the stock using analysis of stock market Twitter data for movement and Twitter predicting stock market movements 8 Sentiment V. S. Pagolu 2021 According to the semi- Predictability for strong form of the Stocks efficient market hypothesis (EMH), markets are efficient and stock prices already reflect all publicly available information. This study focuses on the application of several time-series and NLP techniques to extract data regarding the price of a specific stock. 9 Prosky, 2022 Using a structured Machine Jordan literature review Learning Stock approach, relevant Market peer-revested journal Prediction publications spanning Studies: Review the last 20 years are and Research located. The studies Directions with comparable methodology and settings are categorised. Each study in each category is looked at to look for trends, odd results, limitations, and areas that need more investigation. 10 Short-term stock Strader, 2019 A team of researchers market price Troy J gathered two years' trend prediction worth of data from the using Chinese stock market a comprehensive and offered a thorough deep learning customisation of system feature engineering and deep learning- based models. The method is extensive and includes two different algorithms and data pre- processing to recommend the finest investment sections. 11 A Survey on Shen, Jingyi 2021 In financial markets, Machine machine learning (ML) Learning for has become a powerful Stock Price analytical tool used to Prediction: help and manage Algorithms and investment efficiently. Techniques This paper revested and compared the state-of-the-art of ML algorithms and techniques that have been used in finance, especially the stock price prediction. 12 STOCK MARKET Obthong, 2022 Support Vector PREDICTION Mehtabhorn Machines (SVM) is a SYSTEM USING relatively sophisticated MACHINE learning technique that LEARNING controls the decision APPROACH function, employs the kernel method, and produces sparse solutions. Back propagation neural network method seeks to calculate the learning rate and training cycle parameters in an adaptable manner. The findings indicate that SVM is a potent tool for stock forecasting in the financial sector. 13 Stock Market Ajinkya 2022 This study will show Price Prediction Rajkar how recurrent neural and Analysis networks can be used to predict a stock's closing price. The proposed model can provide buy or sell suggestions by fusing the two methods. Django and React have been used to create an app that implements the suggested system in real-time. 14 Stock Market Deshmukh 2021 In this project, machine Prediction Using learning is being used Machine to predict many Learning features of a specific stock or index, including future values of the opening price, closing price, index value, etc. It aimed to grasp the informational landscape of the financial market and pinpoint the factors that affect stock prices. 15 Stock Market Chouhan 2020 The goal of stock Prediction Using market prediction is to Machine forecast the value of a Learning company's financial stocks in the future. Multiple models are used by machine learning itself to facilitate and authenticate prediction. The paper focuses on LSTM-based machine learning and regression for stock value prediction. 16 A Survey on Polamur 2021 The purpose of this Stock Market study is to discuss Prediction Using various methods for Machine stock market Learning forecasting. The most Techniques recent stock market prediction methods, including Artificial Neural Networks, Neuro-Fuzzy Systems, Time Series Linear Models (TSLM), and Recurrent Neural Networks (RNN) are examined. 17 ROLE OF Awad 2020 A key element in MACHINE optimising a company's LEARNING IN profit is the stock PREDICTING value. Due to its non- STOCK PRICES: A linearity and wide LITERATURE range of abrupt swings, SURVEY stock price prediction is a highly difficult assignment. This literature review examines the most recent machine learning and deep learning methods for estimating stock price values. 18 Twitter mood Bollen 2019 Mood-tracking predicts the programmes used to stock market examine the language in daily Twitter feeds. Findings suggest that the inclusion of some public mood dimensions can greatly increase the precision of DJIA predictions. They observe a reduction in the Mean Average Percentage Error of more than 6% and an accuracy of 87.6%. 19 Stock Price ] Choi, 2018 The ARIMA-LSTM Correlation Hyeong Kyu hybrid model predicts Coefficient the correlation Prediction with between two different ARIMA-LSTM stocks. It outperformed Hybrid Model all other financial models on a meaningful scale. The Constant Correlation model, which in their experiment outperformed the other models, had values that almost halved. 20 A Hierarchical ] Kim, 2019 Their HATS technique Graph Attention Raehyun carefully gathers data Network for on various relation Stock Movement kinds and adds the data Prediction to each company's representations. The approach is utilised to forecast changes in market indices as well as individual stock prices. 21 Forecasting Pushpendu 2020 They use both random directional forests and LSTM movements of networks as training stock prices for approaches. They intraday trading compare how they using LSTM and predict directional out- random forests of-sample changes in the stocks that make up the S&P 500 during intraday trading from January 1993 to December 2018. They outperform the single- feature setting in Fischer & Krauss (2018) and Krauss et al. (2017). 22 Temporal Fuli 2019 Stock prediction Relational attempts to forecast Ranking for the future trends of a Stock Prediction stock. Relational Stock Ranking (RSR) is a fresh deep learning approach for stock prediction. They back-test their methodology using historical data from the NYSE and NASDAQ stock exchanges. It achieves an average return ratios of 98% and 71% on these two markets, respectively. 23 Artificial Hadi 2019 In this paper, Support Counselor Vector Regression is System for Stock used to forecast the Investment future stock prices (technical characteristics). The forecasted prices are then used to suggest which sections of the investor's budget to allocate to various existing equities. The efficiency of the suggested system is demonstrated by the experimental findings on the New York Stock Exchange. 24 Stock Movement Yumo 2018 The market is very Prediction from stochastic, and they Tweets and construct temporally- Historical Prices dependent predictions from chaotic data. They provide a deep generative model that works on this task by combining text and price signals. They put their model to the test on a brand-new, extensive dataset and found that it outperforms strong baselines. 25 Leveraging Huicheng 2018 Financial news often Financial News contains useful data on for Stock Trend the market and publicly Prediction with traded corporations. By Attention-Based using deep neural Recurrent models to extract Neural Network valuable semantic information from news material, they are able to solve this problem. 26 Predicting the Dev 2019 The authors examined Effects of News the stock market's Sentiments on reactions to news the Stock sentiments. They are Market able to anticipate patterns in short-term stock price movement with a directional accuracy of 70.59% using simply news feelings. Use of Twitter data to forecast stock values based on popular opinion has gained traction. 27 Using Deep Kusuma 2019 A study investigates the Learning Neural predictability of the Networks and stock market. Candlestick Candlestick charts and Chart deep convolutional Representation networks are used for to Predict Stock this work. The results Market are used to create a decision support framework that traders can use to offer indicators for the direction of the stock price going forward. 28 Global Stock y.Koh 2018 Their model generates Market profits not just in the Prediction Based stock market of the on Stock Chart nation in which it was Images Using taught, but also Deep Q-Network typically in international stock markets. Findings show that future stock values may be anticipated even when the training and testing processes are carried out in several nations. 29 Applications of Jiang 2019 Stock market deep learning in forecasting has drawn stock market interest from both prediction: economists and recent progress computer scientists. Deep learning models have recently been proposed as new horizons for this subject. This survey aims to provide an evaluation of recent efforts on deep learning models for stock market prediction. 30 Effectiveness of Mokhtari 2021 Using artificial Artificial intelligence (AI), this Intelligence in study examines the Stock Market effectiveness of ML Prediction Based systems for stock on Machine market forecasting. Learning Results demonstrate that it is premature to assert that AI can given current AI capabilities, outperform stock markets. The findings suggest that while AI can forecast public opinion regarding the financial markets, its accuracy is not sufficient. 31 Novel Stock R.Mohan 2022 A stock market crash is Crisis Prediction a drop in stock prices Technique—A more than 10% across Study on Indian the major indices. They Stock Market proposed a crisis prediction model based on the Hybrid Feature Selection (HFS) technique. The effectiveness of the model is quantified by using the XGBoost and DNN regression method. 32-- Global Stock R.Kim 2021 They applied Deep Q- Market Network with a Prediction Based Convolutional Neural on Stock Chart Network function Images Using approximator to make Deep Q-Network global stock market predictions. Their model takes chart images of individual company as input and chooses one action among Long, Neutral, or Short every day. 33 ML-GAT:A F.Liu 2022 A team of researchers Multilevel Graph has proposed a novel Attention Model multilevel graph for Stock attention network (ML- Prediction GAT) for predicting the stock market trends. Their model's linkages are interpretable and consistent with real- world interactions. The results also demonstrate the importance of using financial news and relational data. 34-- Decision Fusion C. 2021 A forecasting model for Stock Market zhang that applies decision Prediction: A fusion can consist of Systematic only homogeneous Review base learners, such as ANN, decision trees, SVM, and LSTM. Data type of forecasts is a critical factor in selecting fusion methods. In general, voting and tree-based methods are the two most common fusion methods for classification. 35 Stock Ranking J.Gao 2022 Stock ranking Prediction Using performance can be List-Wise improved by Approach and incorporating the stock Node relation information in Embedding the prediction task. Technique They employ a graph- based approach for stock ranking prediction. NRBO-k- based investment strategy generates 0.281% to 4.928% higher relative gain than the topmost stock-based strategy. 36 Integrated Long- Yuan 2021 The classical linear Term Stock multi-factor stock Selection selection model is Models Based widely used for long- on Feature term stock price trend Selection and prediction. In this Machine paper, the features are Learning selected by various Algorithms for feature selection China Stock algorithms. Annualized Market return from 2011 to 2018 for the new long- short portfolio is 21.92% while max drawdown is 13.58%. 37-- Ensemble S.Alotaibi 2020 The forecast of the Technique With stock price attempts to Optimal Feature assess the potential Selection for movement of the Saudi Stock financial exchange's Market stock value. The exact Prediction: A estimation of the Novel Hybrid movement of share Red Deer-Grey price would contribute Algorithm more to investors' profit. This paper introduces a new hybrid model referred to Red Deer Adopted Wolf Algorithm (RDAWA). 38 Stock Volatility Hui and Guo 2019 HTPNN captures the Prediction by potential law of stock Hybrid Neural price fluctuation by Network learning the fusion feature of news and time series. Compared with the state-of-the- art methods, their method combines more abundant stock characteristics and advantages in running speed. Besides, the accuracy is averagely improved by nearly 5%. 39-- Predicting Stock Nabipour 2020 This study compares Market Trends nine machine learning Using Machine models (Decision Tree, Learning and Random Forest, Deep Learning Adaptive Boosting) and Algorithms Via two powerful deep Continuous and learning methods Binary Data; a (Recurrent Neural Comparative Network (RNN) and Analysis Long short-term memory (LSTM). Diverse financials, petroleum, non- metallic minerals and basic metals from Tehran stock exchange are chosen for experimental evaluations. 40 A Novel N.Liu 2017 A novel instantaneous Instantaneous frequency called - Frequency counting instantaneous Algorithm and frequency is proposed. Its Application in The newly defined IF Stock Index satisfies the three Movement requirements for Prediction instantaneous frequency. It can be used to simple wave, including IMFs obtained from an EMD algorithm. An experiment is provided based on the stock market data of Hong Kong Hang Seng Index. 41-- Stock Ranking S.Wang 2022 An adversarial game Prediction Based neural network model on an can eliminate the Adversarial influence of market- Game Neural style factors on stock Network ranking predictions. This model involves a mutual game between the main neural network and the auxiliary neural network. It can strengthen the stock selection performance of the model, but it has shortcomings. 42 Stock Market Zhang 2020 Researchers develop a Prediction via multi- Source multiple Multi-Source instance model that Multiple can effectively combine Instance events, sentiments, as Learning well as historical quantitative data into a comprehensive framework. Evaluations on the data from the year 2015 and 2016 demonstrate the effectiveness of their model. Their approach is able to automatically determine the importance of each data Source. 43-- How to Handle D.Akshir 2021 Stock market Data Imbalance forecasting aims to and Feature predict possible future Selection prices or directions of Problems in an index/stock. In CNN-Based literature, the Stock Price convolutional neural Forecasting networks (CNN) models are used for stock market forecasting and gave successful results. But, data imbalance due to labelling and feature selection problems. They are seen when considering these models. 44 A Hybrid E.KOO 2018 Distribution of volatility Prediction time-series is highly Model concentrated near Integrating zero. Such aspect can GARCH Models cause low prediction With a performance on the Distribution whole domain of Manipulation probability density Strategy Based function. Authors on LSTM propose a new hybrid Networks for model with GARCH- Stock Market type models based on a Volatility novel non-linear filtering method to mitigate concentration property of volatility. 45 Predicting the W.Chanag 2022 This study aims to Direction of US predict the direction of Stock Prices US stock prices by Using Effective integrating time- Transfer Entropy varying effective and Machine transfer entropy (ETE) Learning and various machine Techniques learning algorithms. They suggest utilizing the adjusted accuracy derived from the risk- adjusted return in finance as a prediction performance measure. 46-- ST-Trader: A K.Wang 2021 Stocks that are Spatial-Temporal fundamentally Deep Neural connected with each Network for other tend to move Modeling Stock together. Considering Market such common trends is Movement believed to benefit stock movement forecasting tasks. They propose a framework that incorporates the inter-connection of firms to forecast stock prices. Experiments on minute-level U.S. stock market data demonstrate that their model effectively captures both spatial and temporal signals. 47 Forward s.Chou 2020 This paper proposes an Forecast of intelligent time series Stock Price Using prediction system that Sliding-Window uses sliding window Metaheuristic- metaheuristic Optimized optimization for the Machine- purpose of predicting Learning the stock prices of Regression Taiwan construction companies. 48 Deep Learning Lien Minh 2021 This paper proposes a Approach for novel framework to Short-Term predict the directions Stock Trends of stock prices by using Prediction Based both financial news and on Two-Stream sentiment dictionary. Gated Recurrent Results show that two- Unit Network stream GRU outperforms state-of- the-art models. This framework can also be integrated into an automated system to support investors in trading specific stocks. 49 Fast Training J . Wang 2022 Deep convolutional Algorithms for fuzzy system (DCFS) on Deep a high-dimensional Convolutional input space is a Fuzzy Systems multilayer connection With Application of many low- to Stock Index dimensional fuzzy Prediction systems. They apply the DCFS models to predict a synthetic chaotic plus random time-series and the real Hang Seng Index of the Hong Kong stock market. 50-- Stock Prediction s. chen 2021 This paper proposes to Based on use Genetic Genetic Algorithm(GA) for Algorithm feature selection and Feature develop an optimized Selection and Long Short-Term Long Short-Term Memory(LSTM) neural Memory Neural network stock Network prediction model. Although the model proposed in this paper can effectively improve the prediction accuracy and has strong robustness, there are still some shortcomings.