System and method for generating machine learning model with trace data
AssigneeSamsung Display Co., Ltd.
InventorJanghwan Lee
Life of this patent
Estimating the term…
Abstract
A method for detecting a fault includes: receiving a plurality of time-series sensor data obtained in one or more manufacturing processes of an electronic device; arranging the plurality of time-series sensor data in a two-dimensional (2D) data array; providing the 2D data array to a convolutional neural network model; identifying a pattern in the 2D data array that correlates to a fault condition using the convolutional neural network model; providing a fault indicator of the fault condition in the one or more manufacturing processes of the electronic device; and determining that the electronic device includes a fault based on the fault indicator. The 2D data array has a dimension of an input data to the convolutional neural network model.
Claims20
A method for training a machine learning model using time-series data, the method comprising:
- receiving first time-series sensor data having a first size and a first ordering;
- arranging the first time-series sensor data in a two-dimensional (2D) data array, wherein the arranging includes:
- processing the first time-series sensor data to have a second ordering different from the first ordering by:
- determining a first subset of the first time-series sensor data that correlates to a fault condition in an electronic device;
- and
- determining a second subset of the first time-series sensor data that correlates to the fault condition, the first subset being different than the second subset;
- and
- storing the first time-series sensor data in the 2D data array according to the second ordering by rearranging the 2D data array to place the first subset closer to the second subset, such that the fault condition is recognizable;
- and
- training the machine learning model using the 2D data array to correct the fault condition.
A system comprising a processor and a memory storing instructions, wherein the instructions, when executed by the processor, cause the processor to perform:
- receiving first time-series sensor data having a first size and a first ordering;
- arranging the first time-series sensor data in a two-dimensional (2D) data array, wherein the arranging includes:
- processing the first time-series sensor data to have at least one of a second size different from the first size, or to have a second ordering different from the first ordering by:
- placing the first time-series sensor data into a raw trace data array having a first dimension corresponding to a number of sensors and a second dimension corresponding to a number of data samples for each of the sensors;
- and
- separating a first portion of the first time-series sensor data from a second portion of the first time-series sensor data to fit the first portion and the second portion in the 2D data array, based on the 2D data array having a third dimension that is different from at least one of the first dimension or the second dimension;
- and
- storing the first time-series sensor data in the 2D data array according to at least the second size or the second ordering;
- training a machine learning model using the 2D data array;
- and
- correcting a fault condition in an electronic device based on an output of the machine learning model.
A method for training a machine learning model, the method comprising:
- receiving time-series sensor data having a first size and a first ordering;
- arranging the time-series sensor data in a two-dimensional (2D) data array, wherein the arranging includes:
- processing the time-series sensor data to have a second size different from the first size by:
- determining a first number of data samples associated with a first step of a first process;
- determining a second number of data samples associated with a first step of a second process, the second number of data samples being different from the first number of data samples;
- and
- changing the second number of data samples to equal the first number of data samples;
- and
- storing the time-series sensor data in a data array according to the second size, the second size corresponding to the second number of data samples being equal to the first number of data samples;
- training the machine learning model using the 2D data array;
- and
- correcting a fault condition in an electronic device based on an output of the machine learning model.
Dependencies are inferred from the claim text.