ABOUT THE BOOK :

The book,“Development of a Short-term Forecasting model using Extreme Learning Machine” deals with the short-term prediction of stock indices by employing novel machine learning approaches in the field of financial engineering, efficient short-term prediction of various stock indices plays a vital role for the investors. To achieve this objective and to familiarise the new learners in this promising field. The book is written on this theme in a simplified but technically sound manner. The book is organized by embodying six chapters. These are: I. Introduction II. Overview of Stock Market Prediction III. Literature Survey IV. Details of stock exchange data V. Development of ANN based Prediction Models VI. Simulation based experiments VII. Conclusion and Scope for future work The introduction chapter presents the background, motivation of the book and chapter wise contribution. The second chapter deals with the fundamentals of stock market and the need of its prediction. The third chapter covers the relevant update literature on this topic which will facilitate the readers to carry out further research on this field. An overview of stock indices data[Bombay Stock Exchange (BSE) and National Stock Exchange (NSE)] used for simulation study is provided in chapter IV. The development of stock indices prediction models: Back propagation neural network (BPNN), Functional link artificial neural network (FLANN) and most importantly the Extreme learning machine (ELM) are presented in chapter V. In chapter VI the details of simulation-based experiments and the various key results on stock market prediction are dealt. Finally, chapter VII outlines the main contributions and scope for further research work on this field. Objectives The main objectives of the book are to develop, train and validate an ELM based short term prediction of BSE and NSE by using standard stock market datasets. The proposed model is suggested and it is demonstrated that the ELM stock market prediction model outperforms the other two competitive BPNN and FLANN models.Appropriate features have been extracted from the raw stock market data and then fed to the ELM and other two models for training purpose. The unused feature sets are used for testing purpose. In terms of root mean square error (RMSE) and mean absolute percentage error (MAPE), the three models are ranked. It is shown that the proposed ELM model exhibits the best performance. The BPNN and the FLANN models stand in second and third rank.

ABOUT THE AUTHOR :

Prof. (Dr.) Debahuti Mishrais currently a Professor and Head of the Department of Computer Science and Engineering at Faculty Engineering and Technology (Institute of Technical Education and Research) under Siksha ‘O’ Anusandhan (Deemed to be) University, Bhubaneswar, Odisha. She has approximately 27 years of experience in teaching and research. Prof. Mishra has contributed seven books of International and National level on several core subjects and research domain in Computer Science. She also has around 200 research publications to her credit in several International and National Journals, Conferences of repute. Around 150 publications of her are indexed in Scopus database and overall h-index is 12. Most acclaimed research work of Prof. Mishra includes two IEEE transactions on Industrial Informatics, around six journal articles published with impact factor above 1o and around 25 numbers of SCI/SCIE/ESCI research articles. She has successfully supervised sixteen Ph.D. scholars so far and presently fourteen more scholars are under her supervision under Siksha ‘O’ Anusandhan (Deemed to be) University. She serves as the Editor in several journals of National and International repute. She also has organized more than 25 International/National conferences, workshops, symposiums, winter schools etc. Prof. Mishra has a wider domain of research which includes health and bio-informatics, financial market data analysis, medical image processing, agriculture and crop prediction, meta-heuristic optimization, cloud containerization and defence signal transmission etc. The diversified societal contributions of her work include gene expression data analysis, protein structure prediction, gene network discovery and cholesterol motif finding, forecasting of Sensex, mutual fund, gold, crude oil, currency exchange, commodity market analysis, brain disease diagnosis through MRI, agricultural forecast and yield prediction based on climatic changes, software defined radio for defence data transmission etc.

Adyasha Rath  Adyasha Rath is currently a senior PhD student of Computer Science & Engineering department at (Institute of Technical Education and Research) under Siksha ‘O’ Anusandhan (Deemed to be) University, Bhubaneswar, Odisha, India. She is working in the area of detection of Cardiovascular Disease (CVD) using machine and deep learning based method. She has already published two SCI journal papers in the area of CVD. In addition, she has communicated three journal articles for possible publications. She has presented six international conference papers and two book chaters in the area of detection of diseases using machine and deep learning methods. She has consistently good academic record and has topped in her university in her M.Tech batch. Her PhD work will be completed by July 2022. She is a regular reviewer of many international journals. Her specialization is in the area of data analytics, soft computing, evolutionary computing, machine and deep learning techniques and their applications to healthcare, finance and network security.

CONTENTS :

1. Introduction

2. Overview of Stock Market Prediction

3. Literature Survey

4. Details of Stock Exchange Data

5. Development of ANN Based Prediction Models

6. Simulation Based Experiments

7. Conclusion and Scope for Future Work