Feature selection based on mutual information for human activity recognition

Fish, Benjamin; Khan, Ammar Hussain; Chehade, Nabil Hajj; Chien, Chieh; Pottie, Gregory J.

doi:10.1109/icassp.2012.6288232

Cited by 21 publications

(10 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To resolve these issues we use the following quantities: FAR = Number of normal subsequences detected as stumbles Total number of stumbles (6) Precision = Number of stumble subsequences detected as stumbles Number of all subsequences detected as stumbles (7) Recall = Number of stumble subsequences detected as stumbles Number of stumble subsequences (8) In Figure 5, we present the performance of our system as a ROC curve, which is probability of detection versus false alarm rate. We evaluate each of the 7 locations, and show each as a curve.…”

Section: Resultsmentioning

confidence: 99%

“…This method is obviously flawed and not reliable since it relies entirely on the ability to remember and report stumble events. However, the advancements in low cost and low power sensing technology, low cost storage systems, processing systems, and mathematical tools are now enabling us to continuously and remotely monitor the activities of people [1], [6], [16], [20]. In this work, we design a system that monitors the walking of people, and detects stumbles using a single low cost and low power accelerometer.…”

Section: Introductionmentioning

confidence: 98%

See 1 more Smart Citation

Detecting stumbles with a single accelerometer

Chehade

Özişik

Gomez

et al. 2012

2012 Annual International Conference of the IEEE Engineering in Medicine and Biology Society

View full text Add to dashboard Cite

Falls are a common problem in the elderly population, and their prediction has been a major interest for the medical field. The relationship between stumbles and falls has not been very well understood yet. A critical requirement in advancing the study of this relationship is the realization of a realistic and effective stumble detection system. In this paper, we present a system for the detection of stumbles during walking. Our system consists of a single low cost triaxial accelerometer that may be worn by patients and is convenient for a wide range of subjects. We formulate the problem as an anomaly detection and we validate our system with a large data set collected from 9 subjects. The data set contains a total of 100 stumbles and 45 minutes of walking. We compare 7 different placements for the accelerometer, and show that our system achieves a 99% detection rate, with a 0.2% false alarm rate using an accelerometer worn on the chest.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 98%

Detecting stumbles with a single accelerometer

Chehade

Özişik

Gomez

et al. 2012

2012 Annual International Conference of the IEEE Engineering in Medicine and Biology Society

View full text Add to dashboard Cite

show abstract

“…They combined DE with extreme learning machine (ELM) to select the best feature subsets from the original feature set. Fish et al [23] classify 14 activities from 14 tri-axial accelerometer sensors using decision tree. They utilized the combination of filter and wrapper approach based on mutual information to select the most relevant features.…”

Section: Feature Selection Algorithmsmentioning

confidence: 99%

Feature Selection Optimization using Hybrid Relief-f with Self-adaptive Differential Evolution

Zainudin¹,

Sulaiman²,

Mustapha³

et al. 2017

IJIES

View full text Add to dashboard Cite

Abstract:In various classification areas, the curse of dimensionality becomes a major challenge among the researchers. Thus, feature selection plays an important role in overcoming dimensionality problem. Relief-f is one of the filter methods to rank the most significant features based on their relevance. Although relief-f proved to be a powerful technique in filter strategy, but this method only rank the features based on their significant level. Hence, feature selection is embedded to select the most meaningful features based on their rank. Differential evolution (DE) is one of the evolutionary algorithms that are widely used in various classification domains. Simple and powerful in implementation, we combined relief-f with DE in our proposed feature selection method to solving the optimization problem. In this work, population size and generation size were adaptively determined from the number of features from relief-f. The performance of proposed method is compared with several feature selection techniques in order to prove their superiority using ten datasets obtained from UCI machine learning repository.

show abstract

“…They are now implemented into numerous fields including aviation, robotics, gaming, sports and others to measure orientations or directions [1,2,3,4]. Some studies also utilized inertial sensing to classify human activities or reconstruct human motions [5,6,7,8,9,10].…”

Section: Introductionmentioning

confidence: 99%

Non-linear complementary filter based upper limb motion tracking using wearable sensors

Chien

Xia

Santana

et al. 2013

2013 IEEE International Conference on Acoustics, Speech and Signal Processing

Self Cite

View full text Add to dashboard Cite

In this paper, we present a method to reconstruct motion trajectories of the upper body using inertial measurement units (IMUs). We combine the use of complementary filters and biomechanical models to reconstruct upper body motions. At rst, we use complementary lters to combine information from low-frequency part of accelerometers and magnetometers, and high-frequency part of gyros to estimate sensor orientations and gyro bias. Then we use the estimated orientations of the upper arm and forearm to calculate trajectories of upper limb movements. Finally, we determine the set of parameters for complementary lters that minimized training errors. Experimental results indicate higher than 90% accuracy using accelerometers, gyros and magnetometers.

show abstract

Feature selection based on mutual information for human activity recognition

Cited by 21 publications

References 13 publications

Detecting stumbles with a single accelerometer

Detecting stumbles with a single accelerometer

Feature Selection Optimization using Hybrid Relief-f with Self-adaptive Differential Evolution

Non-linear complementary filter based upper limb motion tracking using wearable sensors

Contact Info

Product

Resources

About