A novel feature selection method based on normalized mutual information

Lee, Sungyoung; Park, Young-Tack; d’Auriol, Brian J.

doi:10.1007/s10489-011-0315-y

Cited by 128 publications

(45 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Consequently, it is necessary to select the best features from the extracted ones in order to construct a good feature set. Our proposed method [16] measures the quality of a feature based on two criteria: the relevancy of the feature (or the classification power) and the redundancy of the feature (or the similarity between two selected features). These two criteria are computed from the mutual information of the feature as described in Equations (1) and (3):

Rel (X) = \frac{I (C; X)}{l o g_{2} (false| Ω_{C} false|)}

where X is a feature variable, C is a class variable, and Ω C is the state space of C .…”

Section: The Proposed Systemmentioning

confidence: 99%

Comprehensive Context Recognizer Based on Multimodal Sensors in a Smartphone

Han

Lee²,

Lee³

2012

Sensors

Self Cite

View full text Add to dashboard Cite

Recent developments in smartphones have increased the processing capabilities and equipped these devices with a number of built-in multimodal sensors, including accelerometers, gyroscopes, GPS interfaces, Wi-Fi access, and proximity sensors. Despite the fact that numerous studies have investigated the development of user-context aware applications using smartphones, these applications are currently only able to recognize simple contexts using a single type of sensor. Therefore, in this work, we introduce a comprehensive approach for context aware applications that utilizes the multimodal sensors in smartphones. The proposed system is not only able to recognize different kinds of contexts with high accuracy, but it is also able to optimize the power consumption since power-hungry sensors can be activated or deactivated at appropriate times. Additionally, the system is able to recognize activities wherever the smartphone is on a human's body, even when the user is using the phone to make a phone call, manipulate applications, play games, or listen to music. Furthermore, we also present a novel feature selection algorithm for the accelerometer classification module. The proposed feature selection algorithm helps select good features and eliminates bad features, thereby improving the overall accuracy of the accelerometer classifier. Experimental results show that the proposed system can classify eight activities with an accuracy of 92.43%.

show abstract

Rel (X) = \frac{I (C; X)}{l o g_{2} (false| Ω_{C} false|)}

where X is a feature variable, C is a class variable, and Ω C is the state space of C .…”

Section: The Proposed Systemmentioning

confidence: 99%

Comprehensive Context Recognizer Based on Multimodal Sensors in a Smartphone

Han

Lee²,

Lee³

2012

Sensors

Self Cite

View full text Add to dashboard Cite

show abstract

“…The MRMR and NMI exhibits unbalanced redundancy and relevance, and biased toward the noisier features [14]. Moreover, both of the previous methods require a parameter ( ) to be estimated manually.…”

Section: Proposed Fs Methodsmentioning

confidence: 99%

Multi-scoring feature selection method based on SVM-RFE for prostate cancer diagnosis

Albashish

Sahran

Abdullah

et al. 2015

2015 International Conference on Electrical Engineering and Informatics (ICEEI)

View full text Add to dashboard Cite

Prostate cancer diagnosis is based mainly by microscopic evaluation of prostate tissue biopsy which includes assigning cancer grading. The latter is crucial in evaluating the prognosis or cancer progression and treatment. The common grading system used is Gleason grading system that classifies the prostate cancer into five basic grades based on the architecture and pattern of glandular proliferation. However, this process may be subjected to inter and intra observer variation. Therefore, the main aim of this paper is to develop a computer aided diagnosis (CAD) utilizing supervised machine learning techniques for Gleason grading of prostate histology. The proposed procedure utilizes the main tissue components of the images in an ensemble style to correctly classify the input histopathological image into benign or malignant. Moreover, the texture features of the benign and malignant images can be used to build the proposed ensemble framework. However, not all extracted texture features contribute to the improvement of the classification performance of the proposed ensemble framework. Therefore, to select the more informative features from a set is a critical issue. In this study, a new multi-scoring features selection method based on SVM-RFE and conditional mutual information (CMI) is proposed.

show abstract

“…Therefore, we apply our own feature selection method which measures the feature based on the relevancy and the redundancy from the mutual information of the feature [36]. The relevancy is calculated as in Equation (2):

R e l (X) = \frac{I (C; X)}{l o g_{2} (|Ω_{c}|)}

where X is a feature variable, C is a class variable, and

{normalΩ}_{c}

is the state space of C .…”

Section: System Designmentioning

confidence: 99%

Smartphone Location-Independent Physical Activity Recognition Based on Transportation Natural Vibration Analysis

Hur

Bang

Kim

et al. 2017

Sensors

Self Cite

View full text Add to dashboard Cite

Activity recognition through smartphones has been proposed for a variety of applications. The orientation of the smartphone has a significant effect on the recognition accuracy; thus, researchers generally propose using features invariant to orientation or displacement to achieve this goal. However, those features reduce the capability of the recognition system to differentiate among some specific commuting activities (e.g., bus and subway) that normally involve similar postures. In this work, we recognize those activities by analyzing the vibrations of the vehicle in which the user is traveling. We extract natural vibration features of buses and subways to distinguish between them and address the confusion that can arise because the activities are both static in terms of user movement. We use the gyroscope to fix the accelerometer to the direction of gravity to achieve an orientation-free use of the sensor. We also propose a correction algorithm to increase the accuracy when used in free living conditions and a battery saving algorithm to consume less power without reducing performance. Our experimental results show that the proposed system can adequately recognize each activity, yielding better accuracy in the detection of bus and subway activities than existing methods.

show abstract

A novel feature selection method based on normalized mutual information

Cited by 128 publications

References 27 publications

Comprehensive Context Recognizer Based on Multimodal Sensors in a Smartphone

Comprehensive Context Recognizer Based on Multimodal Sensors in a Smartphone

Multi-scoring feature selection method based on SVM-RFE for prostate cancer diagnosis

Smartphone Location-Independent Physical Activity Recognition Based on Transportation Natural Vibration Analysis

Contact Info

Product

Resources

About