Editor, Editors, USER, admin, Bureaucrats, Check users, dev, editor, Interface administrators, lookupuser, oversight, Push subscription managers, Suppressors, Administrators, translator, Widget editors
17,889
edits
Difference between revisions of "Bilateral Trigeminal neuromotor organic symmetry"
Bilateral Trigeminal neuromotor organic symmetry (view source)
Revision as of 16:06, 8 March 2022
, 2 years ago→Algorithms and network
Gianfranco (talk | contribs) |
Gianfranco (talk | contribs) |
||
| Line 265: | Line 265: | ||
Therefore we first defined the characteristics of the network, defined the appropriate input and desired output (called target into the ANN) of the network. Then we adopted the LM algorithm to train the network as described above. But we used the ANN to test the correlation between the EMG values of the right and left muscles. | Therefore we first defined the characteristics of the network, defined the appropriate input and desired output (called target into the ANN) of the network. Then we adopted the LM algorithm to train the network as described above. But we used the ANN to test the correlation between the EMG values of the right and left muscles. | ||
When the training was completed, we wanted to check and analyze Neural Network Performance. At this aim, we used the mean squared error (MSE) and the coefficient of determination (<math>R^2</math>). The MSE is the most common measurement for evaluation of the dissimilarity between the outcomes of a model such as the ANN and true values. If the MSE value is small, it means that the model can estimate the true value with almost zero errors. The MSE value was computed according to Equation 5:<blockquote> | When the training was completed, we wanted to check and analyze Neural Network Performance. At this aim, we used the mean squared error (MSE) and the coefficient of determination (<math>R^2</math>). The MSE is the most common measurement for evaluation of the dissimilarity between the outcomes of a model such as the ANN and true values. If the MSE value is small, it means that the model can estimate the true value with almost zero errors. The MSE value was computed according to Equation 5: | ||
<blockquote> | |||
<math>MSE=\tfrac{1}{N}\sum_{i=1}^N(CC_{i,e}-CC_{i,t})^2</math> | |||
<math>Eq. 5</math> | |||
</blockquote> | |||
Where, <math>CC_{i,e}</math> is the ''i''th estimated value and <math>CC_{i,t}</math> is the ''<math>i^{th}</math>'' true value. For a more reliable test evaluating the model’s success, the R2 was added as a statistical measure. The most important reason for computing <math>R^2</math> is to obtain a measure of how well upcoming outcomes are likely to be estimated by the model. The <math>R^2</math> value is an indication of the relationship between the outputs and targets. If <math>R^2=1</math>, this indicates that there is an exact linear relationship between ANN outputs and targets. If <math>R^2</math> is close to zero, then there is no linear relationship between outputs and targets. The <math>R^2</math> value was computed according to Equation 6: | |||
<blockquote> | |||
Where, <math>CC_{i,e}</math> is the ''i''th estimated value and <math>CC_{i,t}</math> is the ''<math>i^{th}</math>'' true value. For a more reliable test evaluating the model’s success, the R2 was added as a statistical measure. The most important reason for computing <math>R^2</math> is to obtain a measure of how well upcoming outcomes are likely to be estimated by the model. The <math>R^2</math> value is an indication of the relationship between the outputs and targets. If <math>R^2=1</math>, this indicates that there is an exact linear relationship between ANN outputs and targets. If <math>R^2</math> is close to zero, then there is no linear relationship between outputs and targets. The <math>R^2</math> value was computed according to Equation 6: | |||
< | |||
<math>MSE=\tfrac{\sum_{i=1}^N(CC_{i,e}-CC_{i,t})^2}{\sum_{i=1}^N(CC_{i,e}-\bar{CC}_{e})^2}</math> | |||
<math>Eq. 6</math> | |||
</blockquote> | |||
Where, <math>CC_{i,e}</math> is the ''i''th estimated value, <math>CC_{i,t}</math> is the ''<math>i^{th}</math>'' true value, and ''<math>\bar{CC}_{e}</math>'' is the mean of the estimated values.<br /> | Where, <math>CC_{i,e}</math> is the ''i''th estimated value, <math>CC_{i,t}</math> is the ''<math>i^{th}</math>'' true value, and ''<math>\bar{CC}_{e}</math>'' is the mean of the estimated values.<br /> | ||