Ardi Expression

Visualizations to represent and explain the cognitive process.


The cognitive power of the human brain is what sets humans aside from animals with lower intelligence. With cognition, humans are able to receive, select, store, transform, develop and recall information from external stimuli, therefore, relate to the world and involve in intellectual activities.

Cognitive function is essential to carry out any task from the simplest to the most complex. The human brain then uses visualizations to represent and explain its cognitive process.



Ardi Expression helps to present the problem and solutions through such visualization. It helps users to understand the outcomes from the other functions of the AI.

Ardi Expression includes two major parts: Visualization and Machine Learning Explanation.

It helps model builders to debug the model. It provides transparency to model stakeholders so they can better understand why a particular prediction or decision is made from the model.



Ardi Expression includes following features:

  • Local interpretable model-agnostic explanations
  • Approximate the model to the closest linear model at a local level
  • Take a point and generate several points in vicinity
  • Generate a linear model based on the results of the steps above
  • Calculate the weightage of each input feature

  • Image

    Use Case

    April is an analyst at a real estate company. She uses Ardi to build models to predict rent changes in the market.

    Through the Explanation function, April returns feature attribution values for each test example sent to the model. These attribution values show how much a particular feature affects the rent to the prediction of the baseline, which is the average value of all the features in the training dataset. This technology helps real estate companies with better pricing.


    Learn more about Graphen Ardi

    Request a demo today.