A Buffalo‐based bi‐directional recurrent paradigm for Indian stock market prediction. (7th February 2023)

• Record Type: Journal Article
• Authors: Pattewar, Tareek; Jain, Dinesh; Kiranmayee, B. V.
• Is Part Of: Concurrency and computation. Volume 35:Number 9(2023)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2023-02-07
• Subjects: bidirectional recurrent -- Buffalo fitness -- error rate -- feature selection -- stock; Parallel processing (Electronic computers) -- Periodicals; Parallel computers -- Periodicals; 004.35
• DOI: 10.1002/cpe.7642
• Languages: English
• ISSNs: 1532-0626

Abstract

Summary: Earlier prediction of stock market details is an important factor in finance intersection. Moreover, predicting stock details is difficult as the stock datasets are unstructured and the entities differ based on each application. Hence, the conventional prediction model requires additional features for predicting stock prices. To end these issues, the presented article has focused on developing the novel African Buffalo‐based Bi‐Directional Recurrent Model (ABBRM) to predict stock details efficiently. Once the dataset is collected, preprocessing is performed in the hidden layer of the novel ABBRM. Moreover, the refined data is imported to the classification layer for feature selection and prediction processes. The incorporation of Buffalo fitness has afforded the most satisfactory stock price prediction outcome. Furthermore, the designed procedure is executed in the python environment with appropriate packages. Significant parameters like prediction accuracy, recall, precision, execution time, and error rate verified the projected model's robustness. Moreover, the presented model has attained an average accuracy of 93.5% for the short term and 99.91% for the long term. ABBRM was compared with existing techniques to identify the presented framework's effectiveness and has attained better results than other models. Here, the significance of the novel ABBRM is better prediction, which means the stock features were stored in the Buffalo memory fitness phase; during the execution phases, it helps to attain the exact prediction. Also, the Buffalo algorithm iterations were repeated continuously till a suitable prediction outcome was gained, and better training and testing outcomes were gained by the deep recurrent features in the designed model. Hence, this model is highly suitable for all prediction applications.