선택한 단어 수는 입니다.
최소 단어 이상 선택하여야 합니다.
최소 단어 이상 선택하여야 합니다.
선택한 단어 수는 30입니다.
최대 10 단어까지만 선택 가능합니다.
최대 10 단어까지만 선택 가능합니다.
제안 방법
- 01 ~1. Because the result of learning is different according to the characteristic of problems to which the neural network is applied, however, it is necessary to perform an experiment to find the optimal learning rate for the characteristic of ECG signals. Thus, this study carried out an experiment by varying learning rate from 0.
- Comparing speed and memory of BP algorithm training methods provided by MATLAB6.5, this study selected the best four methods and applied them to ECG data. Here, traincgf is Fletcher Powell Congugate Gradient, trainlm is Levenberg Marquaedt, trainbfg is BFGS Quasi Newton, and trainrp is Resilient Backpropagation.
- In order to classify the pattern of ECG signals, this study used parameters extracted from the characteristics of ECG patterns as input for machine learning. In an experiment with BP algorithm, the use of traincgf resulted in appropriate convergence into a desired point.
- In order to decide wavelet generating function that can remove baseline by minimizing the distortion of raw signals, this study removed baseline by applying various wavelet generating functions.
- In order to extract the characteristics of ST segment, we took as candidate parameters representing the characteristics of ST segment ST0 [amplitude on the starting point of ST segment (R+60 or R+40 ms)], ST60, ST80 [amplitude at R+140 or R+120 ms], the gradient of ST segment, and the area of ST segment (the area surrounded by ST segment and isoelectric level in the interval of ST segment).
- This study applied parameters, which were extracted from the characteristics of electrocardiographic patterns in order to classify the pattern of electrocardiographic signals, to BP algorithm and support vector machine (SVM)[4] and compared the performance of the two models in pattern classification.
- 5GHz PIV and 256M RAM. This study used European ST-T database for experiment, and selected training data and test data as shown in Table 1 and Table 2 respectively. Training data and test data in the 1st to 5th columns are ST0, ST60, ST80, the slope of ST segments and the area of ST segments respectively, and training data in the 6th column indicates if the result of the data is normal (0) or abnormal (1).
※ AI-Helper는 부적절한 답변을 할 수 있습니다.