Create README.txt

Author: celiosantosjr

macrel/data/models/README.txt

@@ -0,0 +1,69 @@
+Load data
+Train RF with OOB estimators
+Train RF without OOB estimators
+Fitting models
+Save AMP model
+Predict test
+# AMP (unbalanced training) CLASSIFIER 
+Accuracy: 0.926
+MCC: 0.862
+
+Confusion matrix:
+[[784 136]
+ [  0 920]]
+
+              precision    recall  f1-score   support
+
+         AMP      1.000     0.852     0.920       920
+        NAMP      0.871     1.000     0.931       920
+
+    accuracy                          0.926      1840
+   macro avg      0.936     0.926     0.926      1840
+weighted avg      0.936     0.926     0.926      1840
+
+
+
+Load bench set
+Train bench set
+Test bench set
+# AMP (benchmark training) CLASSIFIER 
+Accuracy: 0.946
+MCC: 0.893
+
+Confusion matrix:
+[[846  74]
+ [ 26 894]]
+
+              precision    recall  f1-score   support
+
+         AMP      0.970     0.920     0.944       920
+        NAMP      0.924     0.972     0.947       920
+
+    accuracy                          0.946      1840
+   macro avg      0.947     0.946     0.946      1840
+weighted avg      0.947     0.946     0.946      1840
+
+
+
+Load datasets
+Fit Hemo sets
+Save Hemo model
+# Hemo CLASSIFIER 
+Accuracy: 0.936
+MCC: 0.873
+
+Confusion matrix:
+[[102   8]
+ [  6 104]]
+
+              precision    recall  f1-score   support
+
+        Hemo      0.944     0.927     0.936       110
+     NonHemo      0.929     0.945     0.937       110
+
+    accuracy                          0.936       220
+   macro avg      0.937     0.936     0.936       220
+weighted avg      0.937     0.936     0.936       220
+
+January 30th 2023
+scikit-learn 1.1.2