Deep learning-based stock market forecasting: A comparative analysis of ANN, CNN, and LSTM

The financial markets are found to be finite Hilbert space, inside which the stocks are displaying their wave-particle duality. The Reynolds number, an age old fluid mechanics theory, has been redefined in investment finance domain to identify possible explosive moments in the stock exchange. CNX Nifty Index, a known index on the National Stock Exchange of India Ltd., has been put to the test under this situation. The Reynolds number (its financial version) has been predicted, as well as connected with plausible behavioral rationale. While predicting, both econometric and machine-learning approaches have been put into use. The primary objective of this paper is to set up an efficient econophysics’ proxy for stock exchange explosion. The secondary objective of the paper is to predict the Reynolds number for the future. Last but not least, this paper aims to trace back the behavioral links as well.

Until recently, the statistical approach was the main technique in solving the prediction problem. In the framework of static models, the tasks of forecasting, the identification of hidden periodicity in data, analysis of dependencies, risk assessment in decision making, and others are solved. The general disadvantage of statistical models is the complexity of choosing the type of the model and selecting its parameters. Computing intelligence methods, among which artificial neural networks should be considered at first, can serve as alternative to statistical methods. The ability of the neural network to comprehensively process information follows from their ability to generalize and isolate hidden dependencies between input and output data. Significant advantage of neural networks is that they are capable of learning and generalizing the accumulated knowledge. The article proposes a method of neural networks training in solving the problem of prediction of the time series. Most of the predictive tasks of the time series are characterized by high levels of nonlinearity and non-stationary, noisiness, irregular trends, jumps, abnormal emissions. In these conditions, rigid statistical assumptions about the properties of the time series often limit the possibilities of classical forecasting methods. The alternative methods to statistical methods can be the methods of computational intelligence, which include artificial neural networks. The simulation results confirmed that the proposed method of training the neural network can significantly improve the prediction accuracy of the time series.

The proliferation of international business activities drives organizations to expand their operations into new areas and propels human resource management (HRM) to ensure hiring and retaining competent personnel. Consequently, firms have been struggling to place qualified people in relevant roles and provide adequate training. This study utilized information technology to solve these challenges using a web-based system to interconnect the processes, receive the data from the job applicants via a web-based interface, and connect them with suitable employment. Firstly, the proposed model presented a hybrid technique of Convolutional Neural Network (CNN) with Long Short Term Memory (LSTM) Cloud Web-based Human Resource Management System (CLWHRMS) by recognizing distinct features and forecasts the candidate’s potential under various classification tasks. For this, the study used a set of various software tools for web pages and database designing, including for the alteration of images. The hybrid model was executed using real-time data of 250 resumes, which were collected through an online database to validate the overall performance of the developed web-based system in terms of its accuracy, sensitivity, and specificity. Though the specificity was the same with all the techniques, the results illustrated CNN-LSTM technique was 91% accurate and 90% sensitive compared to the traditional methods. This CNN-LSTM model automatically estimates the suitability of a job candidate and projects his/her workability contributing to Saudi Arabian firms to ease and enhance their recruitment process.

Acknowledgment
This project was supported by the Deanship of Scientific Research at Prince Sattam Bin Abdulaziz University, AlKharj, Saudi Arabia under the Specialized Research Grant program with Grant No-2021/02/18747.