MMSE stats, 2 class
Zmienna zależna wyskalowana w niewłaściwy sposób. Zobacz ??? — Marek Grochowski 2015/01/08 19:47
female
Data: mmse1_f_c2_lt8.gmd
- vectors 21
- features 6 (avg, std, kr, sk, min, max from MMSE 1)
- classes: 2 (raven_scaled_pn < 8) 10 (47.6%), 11 (52.4%)
Classification
- GM
- Test : 10 x 5 CV (stratified)
- Transfromation: standarization
| Conf. | Acc [%] | std | note |
|---|---|---|---|
| kNN k=3, Euclid | 69.9 | 18.8 | |
| SVM, Gauss, b=0.22 | 65.0 | 17.3 | 16 SV (almost all training data) |
PCA
MDS
male
Data
- vectors 22
- features 6 (avg, std, kr, sk, min, max from MMSE 1)
- classes: 2 (raven_scaled_pn < 9) 14 (63.6%), 8 (36.4%)
Classification
- GM
- Test : 10 x 5 CV (stratified)
- Transfromation: standarization
| Conf. | Acc [%] | std | note |
|---|---|---|---|
| nic powyżej 63% |
PCA
MDS
