We present a method to detect outlier or exceptional transactions records applying an innovative user modeling. We use a large financial database to validate our method. Our method has two stages. The first stage is for user transaction modeling and it obtains user behavior according to historic transactions based on categorical or numerical attributes. The second stage is the monitoring where a new transaction is compared against the corresponding user model, in order to determine if this transaction is unusual (no standard, fraudulent or suspicious). (e.g. transaction normal, abnormal, suspicious, etc.). And also provides the percentage of ownership to them. According to the experiments conducted with a very large financial database, encouraging results were observed in the field of applied Business Intelligence, in particular to the financial frauds detection and in general to the outlier detection area.
The novelty of this method is that it provides to the user with an automatic explanation about the exception level of the new transaction
Resumen. En la actualidad han surgido nuevos modelos computacionales que intentan superar a los modelos clásicos de optimización, este es el caso de la Computación Evolutiva, la cual se ha popularizado por los Algoritmos Genéticos y sus diferentes variantes que prometen ser mejores. En este artículo analizaremos las bondades y/o deficiencias del Algoritmo Genético básico y del algoritmo de Aprendizaje Incremental Basado en Población, el cual es un algoritmo de estimación de distribuciones que forma parte del paradigma de la Computación Evolutiva. Se presenta un estudio comparativo de ambos algoritmos que permite establecer, a partir de la experimentación realizada con 7 funciones objetivo, que el algoritmo de Aprendizaje Incremental Basado en Población presenta ventaja significativa en tiempo de ejecución de todas las pruebas, así como en la precisión obtenida en 6 de las 7 funciones objetivo analizadas. Aunque esta ventaja ya había sido reportada, en este artículo se ha experimentado con funciones multimodal con dos incógnitas y en tres dimensiones, que en la actualidad son consideradas difíciles de resolver.Palabras clave: Optimización de funciones, computación evolutiva, algoritmos genéticos, algoritmo de estimación de distribuciones, aprendizaje incremental basado en población.
Population-Based Incremental Learning as Good Alternative for Genetic AlgorithmsAbstract. At present, new computational models that attempt to overcome to the classical optimization models have emerged, this is the case of Evolutionary Computation, which has been popularized by Genetic Algorithms and their different variants that promise to be better. In this article we will discuss the benefits and/or shortcomings of basic Genetic Algorithm and Population-Based Incremental Learning algorithm, which is an estimation of distributions
Nowadays, job stress is very common and it has a high cost in terms of workers' health, absenteeism and lower performance. Although stress is not a disease, it is the first sign of a bigger problem which can generate long-term damages. This paper presents a predictive model of job stress which was obtained from data collected by telephone mobile and sensors. Relevant attributes were identified through a correlation analysis. Learning algorithms were applied in order to determine the predictive model. We use the classifier algorithms ZeroR, Naive Bayes, Simple Logistics, Support Vector Machine, k-Nearest-Neighbor, AdaBoost and Random Tree. The proposed model obtained an accuracy of 0.947, a coverage of 0.941 and an F-measure of 0.939. This model was implemented in a mobile application called "TestStress". Also, the results obtained of the experimentation with the app are presented.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.