CPC G06Q 10/087 (2013.01) [G01N 33/146 (2013.01)] | 15 Claims |
1. A computer-implemented method, the method comprising: storing an evaluation for each of a plurality of consumable items in a storage memory, the evaluation of each of the consumable items being expressed in qualitative terms based on a defined panel of metrics and the evaluation of each of the consumable items being linked with a chemistry panel associated with and identifying chemical attributes of each
respective consumable item, each chemistry panel being an attribute vector of raw chemistry data for each respective consumable item obtained by a lab analysis of each respective consumable item, including one or more columns of chemical attributes and a value in a cell of the attribute vector representing a standardized measured value converted from the raw chemistry data; training a machine learning algorithm using the stored evaluation for each of the plurality of consumable items, the machine learning algorithm learning relationships between the qualitative terms and the chemistry panel in the evaluation for each of the plurality of consumable items; executing the trained machine learning algorithm using an untested consumable item having an associated chemistry panel to generate an output including an evaluation of the untested consumable item expressed in the qualitative terms; and storing the output in a memory.
|