Connect HomeMarketplaceComponentsCiphix - Simple Logit Model

Ciphix - Simple Logit Model

Custom Activity

Custom Activity

Updated: September 3rd, 2020

Published: November 12th, 2018

Downloads: 18

Language compatibility: Visual Basic

Community Support

Ciphix - Simple Logit Model

Custom Activity

Summary: 

This machine learning related activity is used to estimate a model for a binary variable and uses that model to make a one point ahead forecast for the binary variable.

The estimated logit model describes a non-linear regression of which the goal is to perform a forecast where the variable we want to predict is either 0 or 1 (binary). The model uses a logistic probability function to determine the relationship between the dependent and the independent variables. The difference in accuracy between this logit model and a linear probability model is shown in the graphs below.

Details

Benefits

This activity enables processes to make predictions for discrete variables, thus incorporating machine learning in the process. The activity can be used to alleviate certain human decisions within a process which can benefit a multitude of automation projects.

Compatibility

Tested with UiPath Studio 2018.2.3

Dependencies

Accord Framework

Resources

- User_Guide_-_SimpleLogit_Ciphix.pdf

Licensing

By clicking download you agree to the following license.