US 12,067,463 B2
Machine learning platform
Stephen Roberts, Abingdon (GB); Mike Osborne, Oxford (GB); Brian Mullins, Oxford (GB); Paul Reader, Oxford (GB); Nathan Korda, Oxford (GB); Rob Williams, Thatcham (GB); and Stanley Speel, Oxford (GB)
Assigned to Mind Foundry Ltd, Oxford (GB)
Filed by Mind Foundry Ltd, Oxford (GB)
Filed on Feb. 18, 2020, as Appl. No. 16/793,814.
Prior Publication US 2021/0256310 A1, Aug. 19, 2021
Int. Cl. G06N 20/00 (2019.01); G06F 18/214 (2023.01)
CPC G06N 20/00 (2019.01) [G06F 18/2148 (2023.01)] 22 Claims
OG exemplary drawing
 
1. A computer-implemented method comprising:
accessing, by one or more processors of a server, a dataset from a datastore;
receiving a definition of a task that identifies a target of a machine learning algorithm from a machine learning platform that operates at the server;
accessing a look up table that maps a type of task with a machine learning tool, the machine learning tool comprising one of a regression tool, a classification tool, and an unsupervised machine learning tool;
identifying, using the look up table, the machine learning tool corresponding to the type of task based on the definition of the task;
forming, utilizing the machine learning algorithm, a machine learning model based on the dataset, the task, and the machine learning tool;
deploying the machine learning model by providing an application that is external to the machine learning platform with access to the machine learning model;
monitoring a performance of the machine learning model after being deployed; and
updating the machine learning model based on the monitoring by:
detecting data deficit based on the performance of the machine learning model, wherein detecting the data deficit comprises identifying missing values and a frequency of the missing values by analyzing statistics of the dataset;
identifying a source of the data deficit as a faulty sensor providing the data;
in response to detecting the data deficit, generating an internal action by accessing without user intervention, from the datastore, additional data that remedy the data deficit, the additional data comprising or replacement data from another dataset of a library of dataset with similar properties to the dataset;
adapting the additional data to match statistical properties of the dataset; and
updating the machine learning model based on the adapted additional data and the task.