当前位置: X-MOL 学术Journal of Risk and Financial Management › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Multi-Layer Perceptron-Based Classification with Application to Outlier Detection in Saudi Arabia Stock Returns
Journal of Risk and Financial Management Pub Date : 2024-02-10 , DOI: 10.3390/jrfm17020069
Khudhayr A. Rashedi 1 , Mohd Tahir Ismail 2 , Sadam Al Al Wadi 3 , Abdeslam Serroukh 4 , Tariq S. Alshammari 1 , Jamil J. Jaber 3, 5
Affiliation  

We aim to detect outliers in the daily stock price indices from the Saudi Arabia stock exchange (Tadawul) with 2026 observations from October 2011 to December 2019 provided by the Saudi Authority for Statistics and the Saudi Central Bank. We apply the Multi-Layer Perceptron (MLP) algorithm for detecting outliers in stock returns. We select the inflation rate (Inflation), oil price (Loil), and repo rate (Repo) as input variables to the MLP architecture. The performance of the MLP is evaluated using standard metrics for binary classification, namely the false positive rate (FP rate), false negative rate (FN rate), F-measure, Matthews correlation coefficient (MCC), accuracy (ACC), and area under the ROC curve (AUC). The results demonstrate the efficiency and good performance of the MLP algorithm based on different criteria tests.

中文翻译:

基于多层感知器的分类及其在沙特阿拉伯股票收益异常值检测中的应用

我们的目标是利用沙特统计局和沙特中央银行提供的 2011 年 10 月至 2019 年 12 月期间的 2026 年观测数据,检测沙特阿拉伯证券交易所 (Tadawul) 每日股价指数中的异常值。我们应用多层感知器(MLP)算法来检测股票收益中的异常值。我们选择通货膨胀率(Inflation)、石油价格(Loil)和回购利率(Repo)作为MLP架构的输入变量。 MLP的性能使用二元分类的标准指标进行评估,即假阳性率(FP率)、假阴性率(FN率)、F-measure、Matthews相关系数(MCC)、准确率(ACC)和面积ROC 曲线(AUC)下。结果表明基于不同标准测试的 MLP 算法的效率和良好性能。
更新日期:2024-02-13
down
wechat
bug