Using OneClassSVM for outliers detection example #1890

Onelinerhub · Dec 16, 2022 · f107d00 · f107d00
1 parent 9cab088
commit f107d00
Showing 1 changed file with 34 additions and 0 deletions.
diff --git a/python-scikit-learn/using-oneclasssvm-for-outliers-detection-example.md b/python-scikit-learn/using-oneclasssvm-for-outliers-detection-example.md
@@ -0,0  1,34 @@
 # Using OneClassSVM for outliers detection example
 
 ```python
 from sklearn import svm
 
 X = [[0.32], [0.31], [0], [0.32], [0.31], [1], [0.32], [0.31]]
 
 m = svm.OneClassSVM(gamma='auto').fit(X)
 detected = m.predict(X)
 ```
 
 - `from sklearn import` - import module from [lib:scikit-learn](https://onelinerhub.com/python-scikit-learn/how-to-install-scikit-learn-using-pip)
 - `X = ` - example values set to detect outliers for
 - `.OneClassSVM(` - creates unsupervised outlier detection model
 - `.fit(` - train transformation model
 - `.predict(` - predict target variable based on given features dataset
 - `detected` - will contain a list of attributed outliers (`-1`) and values that are ok (`1`)
 
 ## Example: 
 ```python
 from sklearn import svm
 
 X = [[0.32], [0.31], [0], [0.32], [0.31], [1], [0.32], [0.31]]
 
 m = svm.OneClassSVM(gamma='auto').fit(X)
 detected = m.predict(X)
 
 print(detected)
 ```
 ```
 [ 1  1 -1  1  1 -1  1  1]
 
 ```