particle swarm optimization algorithm
در نشریات گروه حسابداری-
International Journal of Finance and Managerial Accounting, Volume:10 Issue: 38, Summer 2025, PP 151 -166Fraud increases business risks and costs, creates investor distrust, and questions the professional competence and credibility of accounting. Hence, this study aims to employ data mining methods for fraud risk prediction at the companies listed in the Tehran Stock Exchange within the 2014–21 period. For this purpose, 96 financial ratios were collected by reviewing theoretical foundations and research literature. The proposed classifiers such as the k-nearest neighbors algorithm, Bayesian network, support vector machine, and bagging classifier were adopted for fraud prediction. The performance of all classifiers were evaluated relatively poor . Therefore, financial ratios were reduced to enhance the proposed classifiers through the particle swarm optimization algorithm. In fact, 11 effective financial ratios were extracted with a precision of 72.92% and a prediction accuracy validity of 84,82 %. The extracted ratios were then reevaluated by the proposed classifiers for fraud prediction. According to the reevaluation results, all of the proposed methods improved with the extracted financial ratios. The research results indicated that the bagging classifier yielded the highest precision and accuracy, i.e., 84.28% and 76.85%, respectively, and the lowest prediction error, i.e., 23.15%. It was also 87% efficient in fraud prediction.Keywords: k-nearest neighbors algorithm, Bayesian network, Support Vector Machine, bagging classifier, Particle Swarm Optimization Algorithm
-
بر اساس تحقیقات بازار سرمایه، شوک منفی قیمت سهام در هر بازار تابع عوامل محیطی و ویژگی های خاص شرکتی بوده و هر بینشی درمورد چگونگی تشریح و پیش بینی شوک، می تواند بر تصمیمات سرمایه گذاران و فعالان حاضر در بورس اثرگذار باشد. در این پژوهش بر اساس داده های مرتبط با 140 شرکت ها اقدام به پیش بینی شوک قیمتی سهام با تاکید بر نسبت های مالی شده است. به منظور انتخاب متغیرهای بهینه از مجموعه 96 متغیر، از دو الگوریتم تکاملی بهینه سازی ازدحام ذرات و الگوریتم ژنتیک استفاده شده است. پس از به کارگیری الگوریتم های ذکرشده درنهایت 8 متغیر تاثیرگذار بر شوکهای دایم و موقت استخراج گردید که در مدل رگرسیونی باقیمانده مستحکم در تحقیق تاثیر آنها بر متغیر پیشبینی شونده شوک بررسی گردید. نتایج حاصل از RSME مدلهای بررسی شده بهترتیب برای شوک دایم (الگوریتم ژنتیک)، شوک دایم (الگوریتم تکاملی بهینه سازی ازدحام ذرات)، شوک موقت (الگوریتم ژنتیک) و شوک موقت (الگوریتم تکاملی بهینه سازی ازدحام ذرات)، 5.8433، 5.6284، 7.537 و 7.295 میباشد. همان طور که مشاهده میشود RSME در شوک دایم براساس الگوریتم ژنتیک، بیشتر از RSME مدل شوک دایم براساس الگوریتم تکاملی بهینه سازی ازدحام ذرات میباشد. همچنین در مدل شوک موقت براساس الگوریتم ژنتیک RSME مدل، بیشتر از RSME مدل شوک موقت براساس الگوریتم تکاملی بهینه سازی ازدحام ذرات میباشد. بنابراین می توان بیان نمود که رگرسیون برآورد شده بر اساس متغیرهای انتخابی از الگوریتم تکاملی بهینه سازی ازدحام ذرات دارای RSME پایین تر بوده و قدرت پیش بینی کنندگی بهتری نسبت به متغیرهای انتخابی از الگوریتم ژنتیک دار است.
کلید واژگان: شوک منفی قیمت سهام، رگرسیون، الگوریتم ژنتیک، الگوریتم بهینه سازی ازدحام ذراتAccording to capital market research, the negative stock price shock in any market is a function of environmental factors and specific characteristics of the company, and any insight on how to describe and predict the shock can affect the decisions of investors and activists in the stock market. In this study, based on data related to 140 companies listed on the Tehran Stock Exchange.we have attempted to predict stock price shocks with emphasis on financial ratios. In order to select the optimal variables from the set of 96 variables, two evolutionary algorithms of particle swarm optimization and genetic algorithm have been used. After applying the mentioned algorithms, finally, 8 variables affecting permanent and temporary shocks were extracted, which in the regression model mentioned in the research, their effect on the predictor of shock was investigated. the results of RSME model are the permanent shock (genetic algorithm), permanent shock (particle swarm optimization), temporary shock (genetic algorithm) and temporary shock (particle swarm optimization (particle swarm optimization), 5.8433 , 5.6284 , 7.537 and 7.295 . as we observe , RSME in permanent shock based on genetic algorithm is more than RSME permanent shock model based on the evolutionary algorithm of particle swarm optimization. also in the transient shock model based on the genetic algorithm , the model is more than RSME of the temporary shock model based on the evolutionary algorithm of particle swarm optimization . It can therefore be stated that the estimated regression is based on the selected variables from the evolutionary algorithm of the particle swarm optimization, and has better predictive power than the selected variables of the genetic algorithm.
Keywords: Negative Stock Price Shock, Regression, Genetic Algorithm, particle swarm optimization algorithm -
International Journal of Finance and Managerial Accounting, Volume:8 Issue: 28, Winter 2023, PP 185 -216The stock market involves risks and returns that, if forecasted correctly, can lead to profitability, and for this forecasting, appropriate methods are needed. It is affected by various parameters and needs a way to identify these parameters well and have a dynamic nature. The main goal of this article is forecasting Tehran Price Index (TEPIX) by using hybrid Artificial Neural Network (ANN) based on Genetic Algorithm (GA), Harmony Search (HS) particle Swarm Optimization algorithm (PSO) Moth Flame Optimization (MFO) and Whale Optimization algorithms. GA is used as feature selection. So, PSO, HS MFO and WOA are used to determine the number of input and hidden layers. We use the daily values of the stock price index of the Tehran Stock Exchange from 2013 to 2018 in order to forecasting price and test it. The accuracy of ANN, hybrid Artificial Neural Network with HS, PSO MFO and WOA is evaluated based on different loss functions such as MSE, MAE and etc. the results show that the predictability of Meta-heuristic algorithms in testing period is higher than normal ANN. Also, the predictability of hybrid WOA is higher than hybrid PSO and HS algorithms and MFO.Keywords: Whale Optimization Algorithm, Genetic algorithm, Harmony Search, Particle Swarm Optimization Algorithm, Moth Flame Optimization Algorithm
-
هدف
به باور بنیش (1999)، دستکاری سود زمانی رخ می دهد که مدیریت، اصول پذیرفته شده عمومی حسابداری را به منظور سودآور نشان دادن عملکرد مالی شرکت نقض می کند. در این پژوهش، مدل بنیش با تاکید بر متغیرهای خارج از داده های حسابداری، شامل عدم تقارن اطلاعاتی و بازار رقابت محصول، توسعه یافته است.
روشبرای دستیابی به هدف پژوهش، داده های لازم برای 184شرکت پذیرفته شده در بورس اوراق بهادار تهران، طی سال های 1386 تا 1396 جمع آوری شدند. ضرایب مدل ها، به روش شبکه عصبی آموزش یافته با الگوریتم PSO برآورد شده اند. برای فراهم آوردن قابلیت مقایسه نیز، 10 اجرا با 300 تکرار در هر اجرا انجام گرفت و پس از هم گرایی، اجراها متوقف شدند.
یافته هاتوسعه مدل بنیش، خطای آموزش شبکه عصبی با الگوریتم حرکت تجمعی ذرات را از مقدار 0807/0 به 0777/0 کاهش داد. همچنین، سطح زیرمنحنی راک در مدل بنیش، به ازای بهترین نقطه برش (5021/0) و بهترین دقت (26/60درصد) 5538/0 بود و این سطح در مدل توسعه یافته بنیش به ازای بهترین نقطه برش (5304/0) و بهترین دقت (42/67درصد) به 6335 /0 افزایش یافت.
نتیجه گیرینتایج حاکی از تصادفی بودن مدل بنیش و ناتوانی در تفکیک دو گروه شرکت های دستکاری کننده سود و غیردستکاری کننده سود است. همچنین، نتایج افزایش قدرت مدل توسعه یافته بنیش در قیاس با مدل اصلی را نشان می دهد؛ اما نتیجه آزمون ضعیف است و نشان می دهد که مدل توسعه یافته بنیش نیز در تفکیک دو گروه شرکت های دستکاری کننده سود و غیردستکاری کننده سود، کمابیش یک مدل تصادفی است.
کلید واژگان: الگوریتم حرکت تجمعی ذرات، رقابت در بازار محصول، شبکه عصبی مصنوعی، مدل بنیش، محیط اطلاعاتی شرکتObjectiveAccording to Beneish (1999), “earnings manipulation happens as an instance where management violates Generally Accepted Accounting Principles (GAAP)inordertobeneficiallyrepresentthefirm’s financial performance.” In this research, the development of the Beneish model (DBM) was done through emphasizing non-accounting variables,including the Information Asymmetry (IS) and Product Market Competition (PMC).
MethodsThe data was collected for 184 companies listed in the Tehran Stock Exchange (TSE) during the past 11 years 2006-2017. The coefficients of models were estimated by trained Artificial Neural Network (ANN) through PSO algorithm. In order to provide the potential of comparability, ten run with 300 iterations in each run were done for both the Beneish model (BM) and (DBM), then were stopped after convergence.
ResultsResearch results indicate that training error of ANN trained by PSO algorithm measured by mean square error (MSE) was reduced from 0/0807 to 0/0777 by the development of BM. The area under Receiver Operating Characteristic (ROC) for BM was calculated up to 0/5792, which is located in very low confidence range of 0/5-0/6, indicating failed test result and high prediction error up to 39/74 percent. Consequently the best cut-off point and the best precision for BM were estimated to be 0/5021, 60/26 percent, by the maximum accuracy method, respectively. Furthermore, the results show that the AUC for DBM was increased to 0/6335 through incorporating environmental variables of Product Market Competition (PMC) and information symmetry (IS) to the BM, which is still out of an acceptable range of 0/7–0/8 for a relatively good test, indicating poor test result and high model prediction error up to 32/58 percent. Consequently the best cut-off point and the best precision for DBM were estimated to be 0/5304, 67/42 percent by the intersection point of minimum distance and Youden's index, respectively. Incorporating PMC and IS variables to the original model of Beneish decreased model prediction error from 39/74 to 32/58 percent, which is not statistically significant. Nevertheless, this fact improved the predictive power of the BM slightly insignificant.
ConclusionThe findings indicate that the BM is a random model in Iranian capital market and impotent to detect two groups of earning manipulator and non-earning manipulator companies. Although findings indicate that the DBM is a little bit more powerful than the BM and confirm that the impact of environmental variables of PMC and IS is slightly insignificant, indicating the test outcome is still weak and the DBM is an approximately random model in identifying two groups of earning manipulator and non-earning manipulator companies.
Keywords: Particle Swarm Optimization Algorithm, Product competition market, Artificial Neural Network, Benish model, Information Environment -
Inequality in income distribution is the source of injustices such as class differences based on wealth, incomes, and consumption among members of society. Achieving fair distribution of income requires the proper use of economic instruments among which monetary policy is the most important tool. In this study, using annual data from 1979-2013, the effects of liquidity volume (monetary policy index) on inequality of income distribution (using Gini coefficient inequality index) have been addressed. Therefore, Gini coefficient function was estimated with particle swarm optimization algorithm and genetic algorithm. And based on performance assessment criteria, the model with particle swarm optimization algorithm was chosen to study the impact of monetary policy on income distribution in Iran. Research results show that the relation of liquidity volume variable with direct income distribution and the relation of government expenses with income distribution are significant and indirect. Also, the results show that by increasing human development index, income inequality in society increases and rising inflation reduces inequality of income distribution.Keywords: Monetary policy, Income Distribution, genetic algorithm, particle swarm optimization algorithm, Gini Coefficient
- نتایج بر اساس تاریخ انتشار مرتب شدهاند.
- کلیدواژه مورد نظر شما تنها در فیلد کلیدواژگان مقالات جستجو شدهاست. به منظور حذف نتایج غیر مرتبط، جستجو تنها در مقالات مجلاتی انجام شده که با مجله ماخذ هم موضوع هستند.
- در صورتی که میخواهید جستجو را در همه موضوعات و با شرایط دیگر تکرار کنید به صفحه جستجوی پیشرفته مجلات مراجعه کنید.