Genetic Algorithm (Ga) with Discriminative Learning Based Improved Principal Component Analysis Pca (Dlipca_Ga) for Classification of Polarity Multi-View Textual Data

Abstract

Financial statement misstatement is characterized as “intentional blackmail by a board of directors to harm a financier or lender by deceiving the financial statements.”Even more true is the fact that the amount of fake financial reports filed by certain organizations has increased unimaginably over a decade. To prevent the overwhelming consequences of financial extortion, a strong and reliable strategy for identifying mis statements in financial reports is essential. In this paper, we proposed a hybrid Bayesian Belief Structure Genetic Optimization Algorithm (BBN-GOA) system. This algorithm is widely used to extract and reveal secret insights behind huge amounts of data and plays a key role in extortion detection. The proposed strategy provides better results compared to standard classifiers such as SVM and ANN. Presentation measures are evaluated against various classifiers. The accuracy achieved with this strategy is better than other standard classifiers.