Stock market data is a time-series data in which stock value varies depends on time. Prediction of the stock market is an endeavor to assess the future value of a company’s stock rate which will increase the investor’s profit. The accurate prediction of stock market analysis is still a challenging task. The proposed system predicts stock price of any company mentioned by the user for the next few days. Using the predicted stock price and datasets collected from various sources regarding a certain equity, the overall sentiment of the stock is predicted. The prediction of stock price is done by regression and candlestick pattern detection. The proposed system generates signals on the candlestick graph which allows to predict market movement to a sufficient level of accuracy so that the user is able to judge whether a stock is a ‘Buy/Sell’ and whether to short the stock or go long by delivery. The prediction accuracy of the stock exchange has analyzed and improved to 85% using machine learning algorithms.

Stock market analysis using candlestick regression and market trend prediction (CKRM)

The paper presents a comparative study of the performance of Long Short-Term Memory (LSTM) neural network models with Support Vector Machine (SVM) regression models. The framework built as a part of this study comprises of eight models. In this, 4 models are built using LSTM and 4 models using SVM respectively. Two major datasets are used for this paper. One is the base standard Dow Jones Index (DJI) stock price dataset and another is the combination of this stock price dataset along with external added input parameters of crude oil and gold prices. This comparative study shows the best model in combination with our input dataset. The performance of the models is measured in terms of their Root Mean Squared Error (RMSE), Mean Squared Error (MSE), Mean Absolute Error, Mean Absolute Percentage Error (MAPE) and R squared (R2) score values. The methodologies and the results of the models are discussed and possible enhancements to this work are also provided.

/pdf/a-comparative-study-of-svm-and-lstm-deep-learning-algorithms-13z42t2pku.pdf

A Comparative Study of SVM and LSTM Deep Learning Algorithms for Stock Market Prediction.

Approaches for predicting financial markets, including conventional statistical methods and recent deep learning methods, have been investigated in many studies. However, financial time series data (e.g., daily stock market index) contain noises that prevent stable predictive model learning. Using these noised data in predictions results in performance deterioration and time lag. This study proposes padding-based Fourier transform denoising (P-FTD) that eliminates the noise waveform in the frequency domain of financial time series data and solves the problem of data divergence at both ends when restoring to the original time series. Experiments were conducted to predict the closing prices of S&P500, SSE, and KOSPI by applying data, from which noise was removed by P-FTD, to different deep learning models based on time series. Results show that the combination of the deep learning models and the proposed denoising technique not only outperforms the basic models in terms of predictive performance but also mitigates the time lag problem.

/pdf/forecasting-stock-market-indices-using-padding-based-fourier-2chtjdcmsy.pdf

Forecasting Stock Market Indices Using Padding-Based Fourier Transform Denoising and Time Series Deep Learning Models

This paper is an endeavor to unchain the possibility of forecasting the stock prices of the two major indices of India, i.e., BSE and NSE by adopting econometrics prediction model. For this purpose, the daily closing figures of selected indices were fetched from their respective Web sites. Methodology: The ADF test is applied to detect the presence of unit root and convert the series into stationary. GARCH, ACFs, and PACFs values in correlogram found insignificant at the first difference, and residual of the opted ARIMA model is white noise. ARIMA forecasting model checked the sturdiness of the statistics and found the best forecasting model (0, 1, 4) for DLOG_NSE, and model (3, 1, 2) has chosen for DLOG_BSE from the top twenty best outcomes. Main findings: Having considered 474 observations, the best model for making predictions of the selected indices for the period ranging from 1 March 2018 to 28 March 2018 and the comparison between the predicted and actual performance of both indices exhibited satisfactory outcomes. Applications of this study: The possible insinuation ARIMA model for finding future stock price movements can enable investors to make a square deal of investments because the research-based is on daily closing value rather than emphasizing weekly or monthly data. Novelty/Originality of this study: The paper outlines the best-fitted equation highlighted the strength of the ARIMA model to forecast the daily closing price of the time series data. Therefore, these models are found to be the best predictive model that can be used for forecasting expected value on short-term bases and undeniably can guide investors to select optimum portfolio strategy and varied choice of investment.

Predictive Models for Stock Market Index Using Stochastic Time Series ARIMA Modeling in Emerging Economy

Stock prices are driven by corporate earnings or profit expectations. If a trader thinks that the company's earnings are high or will rise further, they will raise the price of the stock. One way for shareholders to get a return on their investment is to buy low stocks and sell them at high prices. If the company performs poorly and the value of the stock declines, the shareholder will lose some or all of his investment at the time of sale. Therefore, accurate stock price information is important. In this work, we proposed a google stock price prediction model using Recurrent Neural Network (RNN). Previous works on Google stack prediction have used some important techniques and models. Such as deep learning models like Artificial Neural Network (ANN) and Convolutional Neural Network (CNN) has been used for google stock movement prediction. Stock movement prediction on social media data by introducing Stock-Net, Artificial Neural Network (ANN) has also been used with an accuracy score of 0.58. Most of the proposed solutions have limited accuracy. In this paper, we have used Kaggle data of google stock price from the year 2012 to 2016. To predict the stock price of the first two months of 2017 based on the last two months of 2016. For this purpose, we used the Recurrent Neural Network (RNN) as a deep learning model and obtained an accuracy of 87.32%.

Google Stock Prices Prediction Using Deep Learning

In Stock Market Prediction, the aim is to predict the future value of the financial stocks of a company. The recent trend in stock market prediction technologies is the use of machine learning which makes predictions based on the values of current stock market indices by training on their previous values. Machine learning itself employs different models to make prediction easier and authentic. The paper focuses on the use of Regression and LSTM based Machine learning to predict stock values. Factors considered are open, close, low, high and volume.

Stock Market Prediction Using Machine Learning

Twitter is a prevailing online social media platform which is available for a large number of users. On Twitter, users communicate through messages and posts. On it, the messages are termed as tweets. This paper is intended to predict the user’s personality from an analysis of the tweets which the user has shared. There are many techniques for personality prediction but there are some drawbacks of these techniques which are addressed in this paper. The objective of this paper is to provide a general outlook of the measures which are taken to predict the user’s personality and draw comparisons between the results obtained from passing the available data through different classifiers.

Shikhin Gupta

Papers

Stock Market Prediction Using Machine Learning

Personality Prediction and Classification Using Twitter Data