Figure - available from: Soft Computing
This content is subject to copyright. Terms and conditions apply.
Source publication
Stock market trends can be affected by external factors such as public sentiment and political events. The goal of this research is to find whether or not public sentiment and political situation on a given day can affect stock market trends of individual companies or the overall market. For this purpose, the sentiment and situation features are us...
Similar publications
Scientists and regular citizens alike search for ways to manage the widespread effects of the COVID-19 pandemic. While scientists are busy in their labs, other citizens often turn to online sources to report their experiences and concerns and to seek and share knowledge of the virus. The text generated by those users in online social media platform...
Citations
... On the other side, studies that have incorporated Twitter content in their predictions, and at the same time have utilized DL methods like the study by [22] are limited to just a prediction study and not feature interpretation in a nonlinear context. Moreover, it is worth noting that the only mentioned limited works that have attempted to explain or analyze the features through a DL framework have devoted themselves to analyzing the importance of words or sentences in tweets. ...
... However, this is not necessarily witnessed when including the rest of the feature sets. Secondly, while comparing the results of the models that benefit from our proposed feature matrix and the model that has only the price data feature set, across all studied companies and DL architectures, we infer that Twitter data has extractable knowledge that can assist the deep learning models to enhance the estimation of the direction of our sample data, which aligns with the result of the study of [22]. Likewise, by comparing the accuracy of the models that are fed with only embeddings of the text of the tweets (BOW8, BOW16, BOW24, DOC2VEC 8, DOC2VEC 16, DOC2VEC 24) with our proposed feature matrix that instead of embeddings consisted indices such as the sentiments polarity of the text, number of tweets, likes and comments, and the historical influence of the tweet writer we infer that our feature matrix has more effective knowledge to be extracted, which has led to higher accuracy in predictions. ...
Studies conducted on financial market prediction lack a comprehensive feature set that can carry a broad range of contributing factors; therefore, leading to imprecise results. Furthermore, while cooperating with the most recent innovations in explainable AI, studies have not provided an illustrative summary of market-driving factors using this powerful tool. Therefore, in this study, we propose a novel feature matrix that holds a broad range of features including Twitter content and market historical data to perform a binary classification task of one step ahead prediction. The utilization of our proposed feature matrix not only leads to improved prediction accuracy when compared to existing feature representations, but also its combination with explainable AI allows us to introduce a fresh analysis approach regarding the importance of the market-driving factors included. Thanks to the Lime interpretation technique, our interpretation study shows that the volume of tweets is the most important factor included in our feature matrix that drives the market's movements.
... This study [10] investigates the impact of public sentiment and political events on stock market trends using machine learning. Results show sentiment features improve accuracy by 0-3%, and political situation features by about 20%. ...
For many years, academics from various fields have been fascinated by the topic of financial market forecasting. Using machine learning techniques like support vector machines (SVM) and reinforcement learning, several studies have been conducted to anticipate movements in the stock market. The whole body of research on Trading Patterns prediction or trading that used reinforcement learning as their primary machine learning method was thoroughly reviewed. Unreasonable assumptions were made in every article that was evaluated. Therefore, employing a weighted sum unit, the research aims to give a combined Autoregressive Integrated Moving Average (ARIMA) and Long Short-Term Memory (LSTM). The models' performance is maximised on a testing dataset by obtaining and combining predictions from both models with precisely chosen weights. The proposed method aims to capitalize on the complementary strengths of both models, mitigating their individual limitations. The Evaluation metrics such as Root Mean Square Error (RMSE), Mean Absolute Error (MAE), Mean Square Error (MSE) and direction accuracy validate the efficacy
... The various scientific studies that have been carried out on predicting the time series of financial assets are evident [2,[9][10][11][12][13][14]. According to Cho et al. [9], there are two most commonly used ways to predict the future value of these series. ...
... Many works that study financial time series propose the use of different types of data in their machine learning systems. Some studies calculate technical analysis indicators [12], others use technical analysis data in conjunction with fundamental analysis [2,13,14], other studies propose the preprocessing of time series such as the use of lag [10], and others use real and percentage changes [9]. ...
... External factors, such as public sentiment and political events, can also impact stock price fluctuations. This phenomenon was studied by Khan et al. [14], who extracted and processed data from Twitter posts, political news from Wikipedia, and time series from Yahoo! Finance. In the end, Khan et al. [14] applied several machine learning models to the data and concluded that public sentiment data can improve its predictive capacity by up to 3%, while political events generate an improvement of up to 20%. ...
These authors contributed equally to this work. Abstract: Since financial assets on stock exchanges were created, investors have sought to predict their future values. Currently, cryptocurrencies are also seen as assets. Machine learning is increasingly adopted to assist and automate investments. The main objective of this paper is to make daily predictions about the movement direction of financial time series through classification models, financial time series preprocessing methods, and feature selection with genetic algorithms. The target time series are Bitcoin, Ibovespa, and Vale. The methodology of this paper includes the following steps: collecting time series of financial assets; data preprocessing; feature selection with genetic algorithms; and the training and testing of machine learning models. The results were obtained by evaluating the models with the area under the ROC curve metric. For the best prediction models for Bitcoin, Ibovespa, and Vale, values of 0.61, 0.62, and 0.58 were obtained, respectively. In conclusion, the feature selection allowed the improvement of performance in most models, and the input series in the form of percentage variation obtained a good performance, although it was composed of fewer attributes in relation to the other sets tested.
... Predictive features extracted from natural language text are of particular interest in many applications as they can shed light on ideas, beliefs and emotions that influence business decisions. NLP techniques have long been employed to address different practical problems, where they are used to create numerical representations of textual documents, by analysing them in terms of pre-defined lexical categories (Mitra and Gilbert, 2014), sentiment classifications (Desai et al., 2015;Khan et al., 2020), distributed word and paragraph representations (Kaminski and Hopp, 2019), and topic models (Hansen and McMahon, 2016;Thorsrud, 2020;Park et al., 2021), which are subsequently used as input into ML models of various economic phenomena. ...
Reward-Based Crowdfunding offers an opportunity for innovative ventures that would not be supported through traditional financing. A key problem for those seeking funding is understanding which features of a crowdfunding campaign will sway the decisions of a sufficient number of funders. Predictive models of fund-raising campaigns used in combination with Explainable AI methods promise to provide such insights. However, previous work on Explainable AI has largely focused on quantitative structured data. In this study, our aim is to construct explainable models of human decisions based on analysis of natural language text, thus contributing to a fast-growing body of research on the use of Explainable AI for text analytics. We propose a novel method to construct predictions based on text via semantic clustering of sentences, which, compared with traditional methods using individual words and phrases, allows complex meaning contained in the text to be operationalised. Using experimental evaluation, we compare our proposed method to keyword extraction and topic modelling, which have traditionally been used in similar applications. Our results demonstrate that the sentence clustering method produces features with significant predictive power, compared to keyword-based methods and topic models, but which are much easier to interpret for human raters. We furthermore conduct a SHAP analysis of the models incorporating sentence clusters, demonstrating concrete insights into the types of natural language content that influence the outcome of crowdfunding campaigns.
... Stock market forecasting is an especially fascinating theme for the financial benefactors and scientists [6]. Stock market prediction refers to the use of various techniques and models to forecast the future movements of stock prices or the overall market. ...
Owing to the dynamic nature of the financial industry, determining accurate stock market forecasts remains a significant challenge. Traditional forecasting methods often struggle to capture the intricate and volatile dynamics of stock price movements. Similarly numerous strategies for stock market prediction have been presented, precise prediction in this field still requires attention. Based on this insight, a novel sequential training model is proposed by adopting the optimal feature selection procedure. In order to determine stock price predictions, primarily financial Nifty data is obtained from the corresponding source. After acquiring financial data, the feature extraction phase is used to extract features from fundamental analysis, such as the Relative Strength Index, Rate of Change, Average True Range, and Exponential Moving Average. Additionally, statistical characteristics such as mean, standard deviation, variance, skewness, and kurtosis are derived from the stock market data. In order to select parameters, the fitness dependent randomised tunicate swarm optimization technique is utilized after the features have been retrieved. Feature selection improves the deep learning process and increases prediction capability by selecting the most important variables and eliminating irrelevant features. A novel sequential training technique is introduced aimed at forecasting stock market trends by leveraging the chosen features. The suggested approach undergoes comprehensive testing, evaluating its predictive capability using accuracy, precision, and recall metrics, implemented towards enhancing future stock price forecasts.
... Jenkins et al. [8] outlined a novel process for evaluating the human impact of ML applications. Khan et al. [10] studied inclusion of stock market features in ML algorithms and forecast stock prices using various attributes. They have developed this model by edge tweets from Twitter and political news from Wikipedia and further pre-processed the data for discovering the sentiment and situation characteristics for stock market calculation. ...
Nowadays, mobile gaming is one of the interesting issues for today’s generation. It is part of entertainment for many people. Here, we have performed an online survey and collected 203 responses. The analysis is executed through the SPSS 22.0v software. Both unsupervised exploratory analysis and supervised learning is used to deal with the dataset. The result of the study gives emphasis on the human tendency to answer the questions. Here, gender-wise is significant with the sim users, the variety of features in mobile games, time of the playing games, their likings of playing the games, time spent on playing games, types of mobile game that they would like to play. Further, difference between location with types of mobile, mobile game that they would like to play, timings of playing mobile game, time spent on playing mobile games, searching features in mobile games, different types of mobile game that they would like to play, competing with friends to gain the highest score in a game contributes to playing more is observed. The obtained results are helpful in the development of the strategies regarding administrative as well as industry perspective.
... Jenkins et al. [8] outlined a novel process for evaluating the human impact of ML applications. Khan et al. [10] studied inclusion of stock market features in ML algorithms and forecast stock prices using various attributes. They have developed this model by edge tweets from Twitter and political news from Wikipedia and further pre-processed the data for discovering the sentiment and situation characteristics for stock market calculation. ...
In the present study our main focus is on the probable of making forecast of Facebook stock market prices by machine learning algorithms based on historical data. In recent years several tools and methods have been planned and used to check the accuracy of forecasting model. This paper presents a comparative study of machine learning techniques such as Support vector machine (SVM), artificial neural network (ANN) and logistic regression. Further, to optimize the buy or sell strategy we have evaluated cumulative returns and cumulative strategy returns with simulating train test data split. It has been observed that ANN model outperforming as compared to other models. Further, maximum cumulative returns with the cumulative strategy are obtained at split of 90:10. It gives the best performance of the stock with the 32.04% cumulative strategy return. The result suggests that, to optimize the performance of Facebook stock market one can predicted the stock price with proposed strategy in future.
... Machine learning can process large datasets to identify patterns and trends in consumer behavior. Recognizing these trends early can adjust marketing strategies to align with emerging preferences, thus gaining a competitive advantage (Khan et al., 2020). ...
Artificial intelligence (AI) and machine learning technologies are increasingly gaining importance for accurately analyzing consumer behavior and delivering customized advertising content within digital marketing strategies. A rapid increase has been observed in the recent research concerning machine learning and artificial intelligence in the digital marketing literature. This research aims to strategically and thematically present a scientific map of publications using digital marketing, machine learning, and artificial intelligence. For this purpose, the bibliometric analysis method, one of the quantitative research methods, has been used in the study. The data used in this study were obtained from the Scopus database, covering 2007-2023. The gathered data were analyzed and visualized using the Bibliometrix analysis program with its web interface provider, Biblioshiny. A total of 171 publications were reached in the research. There has been an increase in publications since the year 2017. The journal with the most publications is Communications in Computer and Information Science. The most frequently used words are artificial intelligence, marketing, machine learning, commerce, digital marketing, learning systems, machine learning, decision-making, e-learning, and deep learning. Moreover, a co-occurrence network exists between machine learning, marketing, artificial intelligence, digital marketing, and commerce. The highest level of publication collaboration has occurred between the United States of America (USA) and India. The involvement of numerous authors from diverse journals suggests that the topic attracts attention from multiple fields. This study offers an extensive bibliometric analysis to visualize the academic publication landscape of digital marketing, machine learning, and artificial intelligence. Subsequent research is expected to emphasize the real-world applications of these technologies in various industries. The visualization of the data obtained from the analyses in this research holds significance for guiding future studies in this area.
... Additionally, Swathi et al. (2022) [49] showed that sentiment analysis of Twitter data improves the accuracy of price predictions. Khan et al. (2020) [50] found that the prediction of the trends of Google, Microsoft, and Apple stocks using machine learning algorithms is better if it considered the sentiment and political situation. Similar results were obtained by Kaplan et al. (2023) [51], who established that external knowledge and investor sentiment better predict the price movements in the crude oil market. ...
... Additionally, Swathi et al. (2022) [49] showed that sentiment analysis of Twitter data improves the accuracy of price predictions. Khan et al. (2020) [50] found that the prediction of the trends of Google, Microsoft, and Apple stocks using machine learning algorithms is better if it considered the sentiment and political situation. Similar results were obtained by Kaplan et al. (2023) [51], who established that external knowledge and investor sentiment better predict the price movements in the crude oil market. ...
The stock market represents one of the most complex mechanisms in the financial world. It can be seen as a living being with complex ways to enact, interact, evolve, defend, and respond to various stimuli. Technical analysis is one of the most complex techniques based on financial data’s graphical aspects. News sentiment indices are very complex and highlight another important part of behavioral finance. In this study, we propose an integrated approach in order to determine the correlation between news sentiment indices, the stock market, and technical analysis. The research methodology focuses on the stock market’s practical and quantitative aspects. In this sense, we have used the graphical representation of technical analysis and econometric modeling techniques such as VAR and Bayesian VAR. The results of the empirical modeling techniques and analysis reveal some important connections between the stock market and news sentiment indices on the US stock market. The conclusions of this study highlight a strong connection between news sentiment indices, technical analysis, and the stock market which suggests that the behavioral finance aspect is a very important aspect in the analysis of the stock market.
... In the literature, several prediction and classification models are used for time series analysis [2][3][4][5][6][7][8][9][10] and specifically quantitative finance [11,12]. Examples of some statistical and machine learning models that are employed in quantitative finance are autoregressive moving average (ARMA) [13], autoregressive integrated moving average (ARIMA) [14][15][16], seasonal ARIMA (SARIMA) [17,18], exponential smoothing, linear regression, logistic regression [19], least absolute shrinkage and selection operator (LASSO) regression [20,21], Naïve Bayes [22,23], decision tree [24], random forest [22][23][24][25], support vector machine (SVM) [16,19,22,25], k-nearest neighbor (KNN) [23], gradient-boosted decision tree (such as extreme gradient boosting (XGBoost), light gradient-boosting machine (lightGBM)) [15,21,24,26], Prophet [27]. ...
... In the literature, several prediction and classification models are used for time series analysis [2][3][4][5][6][7][8][9][10] and specifically quantitative finance [11,12]. Examples of some statistical and machine learning models that are employed in quantitative finance are autoregressive moving average (ARMA) [13], autoregressive integrated moving average (ARIMA) [14][15][16], seasonal ARIMA (SARIMA) [17,18], exponential smoothing, linear regression, logistic regression [19], least absolute shrinkage and selection operator (LASSO) regression [20,21], Naïve Bayes [22,23], decision tree [24], random forest [22][23][24][25], support vector machine (SVM) [16,19,22,25], k-nearest neighbor (KNN) [23], gradient-boosted decision tree (such as extreme gradient boosting (XGBoost), light gradient-boosting machine (lightGBM)) [15,21,24,26], Prophet [27]. ...
... In the literature, several prediction and classification models are used for time series analysis [2][3][4][5][6][7][8][9][10] and specifically quantitative finance [11,12]. Examples of some statistical and machine learning models that are employed in quantitative finance are autoregressive moving average (ARMA) [13], autoregressive integrated moving average (ARIMA) [14][15][16], seasonal ARIMA (SARIMA) [17,18], exponential smoothing, linear regression, logistic regression [19], least absolute shrinkage and selection operator (LASSO) regression [20,21], Naïve Bayes [22,23], decision tree [24], random forest [22][23][24][25], support vector machine (SVM) [16,19,22,25], k-nearest neighbor (KNN) [23], gradient-boosted decision tree (such as extreme gradient boosting (XGBoost), light gradient-boosting machine (lightGBM)) [15,21,24,26], Prophet [27]. ...
The use of wavelet analysis contributes to better modeling for financial time series in the sense of both frequency and time. In this study, S&P500 and NASDAQ data are separated into several components utilizing multiresolution analysis (MRA). Subsequently, using an appropriate neural network structure,
each component is modeled. In addition, wavelets are used as an activation function in long shortterm memory (LSTM) networks to form a hybrid model. The hybrid model is merged with MRA as
a proposed method in this paper. Four distinct strategies are employed: LSTM, LSTM+MRA, hybrid LSTM-Wavenet, and hybrid LSTM-Wavenet+MRA. Results show that the use of MRA and wavelets as an activation function together reduces the error the most.
