scholarly journals Smooth Support Vector Machine for Suicide-Related Behaviours Prediction

2018 ◽  
Vol 8 (5) ◽  
pp. 3399
Author(s):  
G. Indrawan ◽  
I K P Sudiarsa ◽  
K. Agustini ◽  
Sariyasa Sariyasa
Keyword(s):  
Support Vector Machine ◽  
Psychiatric Hospital ◽  
Medical Records ◽  
Training Data ◽  
Support Vector ◽  
Accuracy Evaluation ◽  
Machine Method ◽  
Support Vector Machine Method ◽  
Further Development ◽  
Smooth Support Vector Machine

Suicide-related behaviours need to be prevented on psychiatric patients. Prediction of those behaviours based on patient medical records would be very useful for the prevention by the psychiatric hospital. This research focused on developing this prediction at the only one psychiatric hospital of Bali Province by using Smooth Support Vector Machine method, as the further development of Support Vector Machine. The method used 30.660 patient medical records from the last five years. Data cleaning gave 2665 relevant data for this research, includes 111 patients that have suicide-related behaviours and under active treatment. Those cleaned data then were transformed into ten predictor variables and a response variable. Splitting training and testing data on those transformed data were done for building and accuracy evaluation of the method model. Based on the experiment, the best average accuracy at 63% can be obtained by using 30% of relevant data as data testing and by using training data which has one-to-one ratio in number between patients that have suicide-related behaviours and patients that have no such behaviours. In the future work, accuracy improvement need to be confirmed by using Reduced Support Vector Machine method, as the further development of Smooth Support Vector Machine.

scholarly journals Separation of Convective and Stratiform Precipitation Using Polarimetric Radar Data with A Support Vector Machine Method

2019 ◽  
Author(s):  
Yadong Wang ◽  
Lin Tang ◽  
Pao-Liang Chang ◽  
Yu-Shuang Tang
Keyword(s):  
Support Vector Machine ◽  
Radar Data ◽  
Training Data ◽  
Precipitation Event ◽  
Support Vector ◽  
Polarimetric Radar ◽  
Machine Method ◽  
Support Vector Machine Method ◽  
Separation Index ◽  
Stratiform Precipitation

Abstract. A precipitation separation approach using support vector machine method was developed and tested on a C-band polarimetric radar located in Taiwan (RCMK). Different from some existing separation methods that require a whole volume radar data, the proposed approach utilizes the polarimetric radar data from the lowest tilt to classify precipitation echoes into either stratiform or convective type. Through a support vector machine method, the inputs of radar reflectivity, differential reflectivity, and the separation index are utilized in the classification. The feature vector and weight vector in the support vector machine were optimized using well-classified training data. The proposed approach was tested with multiple precipitation events including two widespread mixture of stratiform and convective events, a tropical typhoon precipitation event, and a stratiform precipitation event. In the evaluation, the results from the multi-radar-multi-sensor (MRMS) precipitation classification approach were used as the ground truth, and the performances from proposed approach were further compared with the approach using separation index only with different thresholds. It was found that the proposed method can accurately identify the convective cells from stratiform storms with the radar data only from the lowest scanning tilt. It can produce better results than using the separation index only.

scholarly journals A Bayesian least-squares support vector machine method for predicting the remaining useful life of a microwave component

2017 ◽  
Vol 9 (1) ◽  
pp. 168781401668596 ◽  
Author(s):  
Fuqiang Sun ◽  
Xiaoyang Li ◽  
Haitao Liao ◽  
Xiankun Zhang
Keyword(s):  
Support Vector Machine ◽  
Least Squares ◽  
Remaining Useful Life ◽  
Lifetime Prediction ◽  
Support Vector ◽  
Machine Method ◽  
Support Vector Machine Method ◽  
Microwave Component ◽  
Useful Life ◽  
Bayesian Least Squares

Rapid and accurate lifetime prediction of critical components in a system is important to maintaining the system’s reliable operation. To this end, many lifetime prediction methods have been developed to handle various failure-related data collected in different situations. Among these methods, machine learning and Bayesian updating are the most popular ones. In this article, a Bayesian least-squares support vector machine method that combines least-squares support vector machine with Bayesian inference is developed for predicting the remaining useful life of a microwave component. A degradation model describing the change in the component’s power gain over time is developed, and the point and interval remaining useful life estimates are obtained considering a predefined failure threshold. In our case study, the radial basis function neural network approach is also implemented for comparison purposes. The results indicate that the Bayesian least-squares support vector machine method is more precise and stable in predicting the remaining useful life of this type of components.

Sign in / Sign up

Export Citation Format

Share Document