search

rsehgal / TomoML

Application of ML in Muon Tomography
2 stars 0 forks source link

Workout various existing outlier detection algorithms #2

Open rsehgal opened 5 years ago

rsehgal commented 5 years ago

One should find the various existing outlier detection algorithms, All the algorithms name should be listed here, and there results comparison must be reported.

rsehgal commented 5 years ago

One outlier method may be by using ML

rsehgal commented 5 years ago

@apoorvabh98 : Any update on the outlier detection algorithms list ???.

Ani1211999 commented 5 years ago

Official documentation of all 20 algorithms in the pyod library https://pyod.readthedocs.io/en/latest/pyod.models.html Really useful and well explained

rsehgal commented 5 years ago

Hi Apoorva, Please try to apply the outlier detection and removal in 2D instead of 3D. ie. First on XY co-ordinate Then on XZ co-ordinate Then on YZ co-ordinate

And remove outlier in these three planes individually.

Finally only those points will be considered as genuine which are not declared as outlier in all the three plane.

apoorvabh98 commented 5 years ago

3D plot for KNN classifier KNN3D

apoorvabh98 commented 5 years ago

3D plot for random forest RF3D

rsehgal commented 5 years ago

Thanks Apoorva,

You can download the test file from following link https://cernbox.cern.ch/index.php/s/GTx1zi3YUvjfVcC

apoorvabh98 commented 5 years ago

3D plot using knn classifier for the above mentioned test file Figure_1