Quantum Leap Analyst’s patented machine learning technology will automatically detect patterns in data. Patterns define subsets of observations that share similar characteristics and similar outcomes with respect to a chosen target, and the discovery process will detect which attributes are most relevant and informative against that target.
The Collections and Patterns view will filter the pattern collection, retrieve and sort the “top 5, 10 or 20” patterns found in the discovery process based on user defined selection criteria (e.g. data coverage, strength, target state coverage etc.). As mentioned, patterns carry important qualitative and quantitative information for you to judge their relevance. They can be used as filters to drill-down into the associated data subset for further analysis, using the available Exploration features. You can save the pattern as well as the data subset covered by the pattern for other usage.
Attribute Importance is a feature that will help the analyst detect and rank the most important attributes associated with a given target attribute. Attribute Importance is especially valuable as it shows the relative importance of the combination of attributes that comprise a pattern.