Distributed and redundant machine learning quality management

#The present disclosure relates generally to predictive computer models and, more specifically, to the creation and operation of numerous machine learning (and other types of artificial intelligence) models in a form that is modular, reusable, easier to maintain, faster to development, easily adapted and modified by data scientists and analysts alike.

#Data scientists today often struggle to manage predictive models that number in the dozens or hundreds. This is the case because of the ever-increasing appetite for predictive analytics, fueled by the existence of open source tools and libraries. Such computer models are often created or trained on different datasets, by those seeking to provide support and insights for different objectives. Some companies run tens of thousands of propensity models every month just to keep up.

#Not only is the development of machine learning models time consuming, it creates a legacy of technology debt, and the constant drive to improve performance yields to the processing of the datasets is inconsistent from application to applications. The choice of feature management, labeling, selection of positive and negative machine learning classes, decision on which data is duplicated or redundant, which should be abstracted or aggregated varies from implementation to implementation.

#The above-mentioned aspects and other aspects of the present techniques will be better understood when the present application is read in view of the following figures in which like numbers indicate similar or identical elements:

#FIG. 1 is a block logical and physical architecture diagram showing an embodiment of a controller in accordance with some of the present techniques;

#FIG. 2 is a block diagram showing an embodiment of an object-orientation orchestrator in accordance with some of the present techniques;

#While the present techniques are susceptible to various modifications and alternative forms, specific embodiments thereof are shown by way of example in the drawings and will herein be described in detail. The drawings may not be to scale. It should be understood, however, that the drawings and detailed description thereto are not intended to limit the present techniques to the particular form disclosed, but to the contrary, the intention is to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the present techniques as defined by the appended claims.

#To mitigate the problems described herein, the inventors had to both invent solutions and, in some cases just as importantly, recognize problems overlooked (or not yet foreseen) by others in the fields of computer science and data science. Indeed, the inventors wish to emphasize the difficulty of recognizing those problems that are nascent and will become much more apparent in the future should trends in industry continue as the inventors expect. Further, because multiple problems are addressed, it should be understood that some embodiments are problem-specific, and not all embodiments address every problem with traditional systems described herein or provide every benefit described herein. That said, improvements that solve various permutations of these problems are described below.

#FIG. 1 is a schematic block diagram of an example of a controller 10, operating within a computing system 100, in which the present techniques may be implemented. In some embodiments, the computing system 100 may include an entity log repository 12, which in some cases may include entity events 14 and entity attributes 16. Entity events may include targeted actions, non-targeted actions, or both. The computing system 100 may further include a targeted action repository 18 and a plurality of potential targeted actions 20.