利用优化算法和基于松鼠的分类技术通过挑选特征集合来检测欺诈性金融票据
Detection of fraudulent financial papers by picking a collection of characteristics using optimization algorithms and classification techniques based on squirrels
摘要:
为了做出重要的投资决策,投资者需要财务记录和经济信息。然而,大多数公司通过夸大财务报表来操纵投资者和金融机构。欺诈性金融活动存在于任何货币或金融交易场景中,无论是实物交易还是电子交易。在这一领域出现的一个具有挑战性的问题是影响和困扰个人和机构的问题。这一问题在该领域引起了更多的关注,部分原因是金融欺诈的普遍存在和以往研究的缺乏。为此,在本研究中,解决该问题的主要方法是使用基于松鼠优化模式的特征选择和分类方法相结合的基于异常检测的方法。我们的目的是开发这种方法,以提供一个模型,用于检测异常的财务报表中使用的最近邻分类,神经网络,支持向量机,贝叶斯的组合选择的功能。然后分析异常样本,并使用评估标准与推荐的技术进行比较。松鼠优化的元探索能力,以及该方法识别财务数据异常的能力,已被证明是有效的,在实施建议的策略。由于他们的专业知识,他们发现了虚假的财务报表。
英文摘要:
To produce important investment decisions, investors require financial records and economic information. However, most companies manipulate investors and financial institutions by inflating their financial statements. Fraudulent Financial Activities exist in any monetary or financial transaction scenario, whether physical or electronic. A challenging problem that arises in this domain is the issue that affects and troubles individuals and institutions. This problem has attracted more attention in the field in part owing to the prevalence of financial fraud and the paucity of previous research. For this purpose, in this study, the main approach to solve this problem, an anomaly detection-based approach based on a combination of feature selection based on squirrel optimization pattern and classification methods have been used. The aim is to develop this method to provide a model for detecting anomalies in financial statements using a combination of selected features with the nearest neighbor classifications, neural networks, support vector machine, and Bayesian. Anomaly samples are then analyzed and compared to recommended techniques using assessment criteria. Squirrel optimization's meta-exploratory capability, along with the approach's ability to identify abnormalities in financial data, has been shown to be effective in implementing the suggested strategy. They discovered fake financial statements because of their expertise.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。