Two approaches for novelty detection using random forest

Zhou, Qifeng; Zhou, Hao; Ning, Yongpeng; Yang, Fan; Li, Tao

doi:10.1016/j.eswa.2014.12.028

Cited by 22 publications

(6 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In literature, Random Forest is utilized in various financial applications such as: exploring financial soundness of companies (Zhou et al, 2015), predicting stock index movement (Patel et al, 2015), developing seasonal stock trading model (Booth et al, 2014), predicting future performance of companies .…”

Section: End Formentioning

confidence: 99%

Random Regression Forest Model using Technical Analysis Variables

Emir¹,

Dınçer²,

Hacıoğlu³

et al. 2016

IJFBS

View full text Add to dashboard Cite

The purpose of this study is to explore the importance and ranking of technical analysis variables in Turkish banking sector. Random Forest method is used for determining importance scores of inputs for eight banks in Borsa Istanbul. Then two predictive models utilizing Random Forest (RF) and Artificial Neural Networks (ANN) are built for predicting BIST-100 index and bank closing prices. Results of the models are compared by three metrics namely Mean Absolute Error (MAE), Mean Square Error (MSE), Median Absolute Error (MedAE). Findings show that moving average (MAV-100) is the most important variable for both BIST -100 index and bank closing prices. Therefore, investors should follow this technical indicator with respect to Turkish banks. In addition ANN shows better performance for all metrics.

show abstract

Section: End Formentioning

confidence: 99%

Random Regression Forest Model using Technical Analysis Variables

Emir¹,

Dınçer²,

Hacıoğlu³

et al. 2016

IJFBS

View full text Add to dashboard Cite

show abstract

“…As far as we know, there are few works on novelty detection using an ensemble approach. The one of the most outstanding research, using random forest for novelty detection, is proposed by Zhou et al [7]. Zhou et al (2015) made full use of the vote distribution from trees and find a metric to measure the proximity of different samples.…”

Section: Introductionmentioning

confidence: 99%

“…Finally, our approach is appropriate to deal with the high-dimensional data. It should be pointed out that Zhou et al [7] use a vlaue generated from vote information of trees to characterize a class while we use a vector generated from probability information from individual learners to charaterize a class.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Ensemble probability distribution for novelty detection

et al. 2018

View full text Add to dashboard Cite

This paper explores a new ensemble approach called Ensemble Probability Distribution Novelty Detection (EPDND) for novelty detection. The proposed ensemble approach provides a metric to characterize different classes. Experimental results on 4 real-world datasets show that EPDND exhibits competitive overall performance to the other two common novelty detection approaches - Support Vector Domain Description and Gaussian Mixed Models in terms of accuracy, recall and F1 scores in many cases.

show abstract

“…Novel operating scenarios must be identified in order to avoid diagnosis misclassifications and incorrect maintenance scheduling. In this sense, the task of detecting patterns that differs from those available during the training of the monitoring scheme, is called novelty detection [23], [49].…”

Section: Fault Detection and Identification Systemsmentioning

confidence: 99%

Fault detection and identification methodology under an incremental learning framework applied to industrial electromechanical systems

Cariño Corrales

View full text Add to dashboard Cite

Condition Based Maintenance is a program that recommends actions based on the information collected and interpreted through condition monitoring and has become accepted since a decade ago by the industry as a key factor to avoiding expensive unplanned machine stoppages and reaching high production ratios. Among the condition based maintenance strategies, data-driven fault diagnosis methodologies have gained increased attention because of the high performance and widen range of applicability due to less restrictive constrains in comparison to other approaches. Therefore, an increased effort is been made to develop reliable methodologies that could diagnose multiple known faults on a machine with initial applications in controlled environments like laboratory test benches. However, applying those methods to industry applications still represent an ongoing challenge due to the multiple limitations involved and the high reliability and robustness required. One of the most important challenges in the industrial sector refers to the management of unexpected events, in respect of how to detect new faults or anomalies in the machine. In addition, the information initially available of the monitored industrial machine is usually limited to the healthy condition, therefore is not only necessary to detect these new scenarios but also incorporate this information to the initial base knowledge. In this regard, this thesis present a series of complementary methodologies that leads to the implementation of a fault detection and identification system capable to detect multiple faults and new scenarios of industrial electromechanical machines under an incremental learning framework to include the new scenarios detected to the initial base knowledge while achieving a high performance and generalization capabilities. Initially, a methodology to increase the performance of novelty detection models to detect unexpected events in electromechanical system is proposed. Then, a methodology to implement a sequential fault detection and identification system composed by a novelty detection and a fault diagnosis stages with high accuracy is proposed. Finally, two different methodologies are proposed to provide the sequential fault detection and identification system the capacity to include new scenarios to the base knowledge. The proposed methodologies have been validated by means of experimental data of laboratory test benches and industrial electromechanical systems. "Mantenimiento basado en la condición" es un programa que recomienda una serie de medidas preventivas basadas en la información recopilada e interpretada mediante el constante monitoreo de la condición de la maquinaria y ha sido aceptado desde hace una década por la industria como un factor clave para evitar paradas no planificadas de la maquinaria y alcanzar altos índices de producción. Entre las estrategias de mantenimiento basadas en la condición, las metodologías de diagnóstico de fallos basadas en datos han recibido mucha atención debido al alto rendimiento y amplio rango de aplicabilidad, esto se debe que cuentan con menos limitaciones en comparación con otros enfoques. Por lo tanto, se ha hecho un mayor esfuerzo para desarrollar metodologías fiables que puedan diagnosticar múltiples fallos conocidas en una máquina, siendo aplicado inicialmente en entornos controlados como bancadas de laboratorio. Sin embargo, aplicar estos métodos en la industria sigue representando un desafío debido a las múltiples limitaciones implicadas y la alta fiabilidad y robustez requeridas. Uno de los desafíos más importantes en el sector industrial consiste en la gestión de eventos inesperados, específicamente en cómo detectar nuevos fallos o anomalías máquina. Además, la información inicialmente disponible de la máquina industrial monitorizada se limita generalmente al estado sano, por lo tanto, no sólo es necesario detectar estos nuevos escenarios, sino también incorporar esta información al conocimiento base inicial. En este sentido, esta tesis presenta una serie de metodologías complementarias que conducen a la implementación de un sistema de detección e identificación de fallos capaz de detectar múltiples fallos y nuevos escenarios de máquinas electromecánicas industriales en un marco de aprendizaje incremental para incluir los nuevos escenarios detectados al conocimiento base inicial manteniendo un alto rendimiento y capacidades de generalización. Inicialmente, se propone una metodología para aumentar el rendimiento de los modelos de detección de novedad para detectar eventos inesperados en el sistema electromecánico. Después, se propone una metodología para implementar un sistema secuencial de detección e identificación de fallas con alta precisión compuesto por una etapa de detección de novedades y otra de diagnóstico de fallos. Finalmente, se proponen dos metodologías diferentes para proporcionar al sistema secuencial de detección e identificación de fallas la capacidad de incluir nuevos escenarios al conocimiento base. Las metodologías propuestas han sido validadas mediante datos experimentales de bancadas de laboratorio y sistemas electromecánicos industriales.

show abstract

Two approaches for novelty detection using random forest

Cited by 22 publications

References 10 publications

Random Regression Forest Model using Technical Analysis Variables

Random Regression Forest Model using Technical Analysis Variables

Ensemble probability distribution for novelty detection

Fault detection and identification methodology under an incremental learning framework applied to industrial electromechanical systems

Contact Info

Product

Resources

About