I'm struggling to re-implement and catch the results of one of the unsupervised anomaly detections, which are shown below: The credit of picture to this paper
gdt
executable
owl
lottie
near
arules
ktrain
django-piston
tsdoc
palm
google-speech-api
kinect-sdk
hive-table
telerik-grid
xargs
html-rendering
framemaker
microsoft-test-manager
eclipse-mars
clickstream
soft-hyphen
pyaudio
singlethreadmodel
angular-e2e
aws-shell
ternary-search-tree
sysbios
urllib
qplaintextedit
ctypes