IntroductionΒΆ

Intuceo product enables business analysts and other knowledge workers to conduct rapid analytics and predictive studies with an easy-to-use, linear workflow. Intuceo uncovers actionable insights and translates them into easy-to-understand rules and visualizations. Intuceo Analyst is a GUI based offering either on premise or on cloud or as Saas. The core functionality in Intuceo is exposed through a set of REST APIs so that it can be integrated into enterprise business intelligent systems and workflows. This document is primary designed to provide all the information necessary for the enterprise development team working on such integration activities. This covers a brief overview of Intuceo followed by an in-depth description of the APIs.

Intuceo overview

Intuceo allows a user to upload data, start an analysis, clean and preprocess the data so that it is ready to be processed by the core algorithms. When once data is prepared, user may extract insights or build models or do both. User may play with the existing rules (act_on_insight, generalize, specialize or customize) to generate additional rules (insights). Intuceo enables the user to upload evaluation data and make predictions. User can get suggestions so as to convert non-desired class scenarios to desired class by changing the level (value) of actionable attributes. The user may also get best pattern that explains the prediction for that record in the evaluation dataset. The following picture shows the steps in the workflow:

_images/intuceo-overview.png

The following table describes the terms encountered in the API documents

Term Description
Dataset A data file in csv format
Analysis One dataset may be used to answer many questions. An analysis is akin to one project with one specific target attribute (column)
Sub Analysis Minor variations within the scope of one analysis are structured as sub analyses. Every Analysis has at least one sub analysis. All subsequent activities happen on a single sub analysis
Target The attribute whose level (value) forms the crux of the analysis
Attribute A column in the dataset. May be numeric or categorical
Level Value an attribute can assume; a data cell in the input dataset
Bin Intuceo converts all numerical attributes into categorical data by defining Bins. One Bin shall be a range of numerical values
Merge When multiple levels are treated as a single level, they are said to be merged
Evaluation data After a model is built or trained using the initial data set, user can get predictions for a different data file(s). Such a data file is called an evaluation dataset or eval data set in short. User may optionally include predictions generated by some other system in the evaluation data file