| CPC G06F 16/2365 (2019.01) [G06F 16/164 (2019.01); G06F 16/178 (2019.01); G06F 18/2155 (2023.01); G06N 3/044 (2023.01)] | 10 Claims |
|
1. A system for determining whether inconsistencies exist in an entity's shared databases using a machine learning model, said system comprising:
a repository including a plurality of databases that store data and information in a format accessible to users, said plurality of databases being shared by multiple users across multiple divisions and organizations associated the entity;
a back-end server operatively coupled to the repository and being responsive to the data and information from all of the databases, said back-end server including:
at least one processor for processing the data and information;
a communications interface communicatively coupled to the at least one processor; and
a memory device storing data and executable code that, when executed, causes the at least one processor to:
collect data and information from the plurality of databases;
store the collected data and information in the memory device;
process the stored data and information through the machine learning model to determine whether inconsistencies in the data and information exist in and across the plurality of databases including determining if a name of a database is consistent with the data in the database, wherein the machine learning model uses at least one neural network having nodes that have been trained to determine whether inconsistencies in the data and information exist in the plurality of databases,
wherein the machine learning model trains the nodes in the at least one neural network by unsupervised learning and by training nodes in a neural network simulation model that is a simulation of the at least one neural network, said neural network simulation model employing training data to train the nodes in the neural network simulation model for determining whether inconsistencies in the data and information exist in the plurality of databases and exist in a target variable, said target variable being a variable in the neural network simulation model,
wherein the unsupervised learning of the at least one neural network configures the at least one neural network to generate a self-organizing map, to reduce the dimensionally of an input data set, and to perform outlier/anomaly determinations to identify data points in the input data set that fall outside of a normal pattern of the data,
wherein training the nodes in the neural network simulation model uses an iterative training and testing loop that incorporates weights associated with the nodes in the neural network simulation model and iterative calculations that are tested, compared to the target variable and updated in subsequent iterative calculations to improve predictability of the target variable,
wherein training the nodes in the neural network simulation model includes ingesting incoming data by cleaning and transforming new training data into a format that the neural network simulation model can digest,
wherein as newly trained neural network simulation models are generated from new data, preprocessing steps are tied to the newly trained neural network simulation model, and when the preprocessing is updated with newly ingested data, an updated neural network simulation model is generated;
receive a result from the machine learning model of whether inconsistencies in the data and information do exist in the databases;
transmit a communication message through the interface identifying that inconsistencies in the data and information do exist in the databases, where remedial steps could then be taken to remove the inconsistencies from the databases;
rename the databases if the type of data contained therein changes;
analyze and filter the data so that the databases that have the same data are identified and reconciled with each other; and
change the titles of the databases having the same data to be the same title.
|