To search, Click below search items.

 

All Published Papers Search Service

Title

Formulated Dynamic Unsupervised Machine Learning Algorithm Trained over Fused Solitary Value of Key Performance Ratios of Stock for Acquiring Optimized Portfolio

Author

S. M. Khalid Jamal A. A. Salam A. R. Zaki† H. Abdullah

Citation

Vol. 19  No. 8  pp. 82-90

Abstract

One of the most important aspects of financial management is portfolio selection. Over time, its performance analysis is raising its integrity due to the fact that these affect its shareholders, thus enhancing its importance in the global financial market as well. Traditional models like Markovitz mean variance, doesn’t consider the actual financial position of the firm in its computational procedures, which is fairly injustice to the evaluation procedure. Making the presentation a great drawback as the smallest change in the model would enormously affect in decision making. For example, companies may be selected solely on basis of high return while chances of bankruptcy or financial instability being evident, making the evaluation, selection and optimization inappropriate. Contrary to the above, a different approach to achieve efficient evaluation and selection of scripts has been introduced in the current undertaken seminal research. This is achieved by calculating the key performance rations of the firm present for the evaluation of the financial positioning. The evaluated measures will be unified (The calculated values are summed-up to a single value for each script i.e. for individual company) which in turn will be utilized with modified K-Means dynamic clustering algorithm.

Keywords

modified k-means, dynamic clustering, performance evaluation.

URL

http://paper.ijcsns.org/07_book/201908/20190813.pdf