Demo library

## Random Robust Cut Forest - Moody's Analytics

#### Installation

pip install RRCF_Outlier_Detection

#### Objectives

1. Use the RRCTree included in rrcf package as a week learner for creating a forest, incrementing the outlier

detection power.

2. Do the code more user-friendly for its fast implementation

3. Automatize the process of outlier detection through the usage of 3 Sigma analysis

#### Functions

Outlier_Detector ( x, num_trees, num_samples per tree )

• Trains the RRCForest

• Parameters:

x Numpy Array / Data from which we want to detect outliers

num_trees int / Number of trees that are going to be used a weak learners for the forest

num_samples per tree int / Number of samples per tree. this parameter is recommended to be established as (1 / Estimated Proportion of Outliers)

• Attributes:

rrcf_outlier_score () Returns a Pandas' series with the CoDist scores for all input samples

 rrcf_outlier_detector () Returns a Numpy array with the detected outliers from the input samples

#### Citing

M. Bartos, A. Mullapudi, & S. Troutman, rrcf: Implementation of the Robust Random Cut Forest algorithm for anomaly detection on streams, in: Journal of Open Source Software, The Open Journal, Volume 4, Number 35. 2019

## Project details

Uploaded source
Uploaded py3