Methods and mechanisms for preventing fluctuation in machine-learning model performance
An electronic device manufacturing system configured to receive, by a processor, input data reflecting a feature related to a manufacturing process of a substrate. The manufacturing system is further configured to generate a characteristic sequence defining a relationship between at least two manufacturing parameters, and determine a relationship between one or more variables related to the feature and the characteristic sequence. The manufacturing system is further configured to determine a weight based on the determined relationship and apply the weight to the feature. The manufacturing system is further configured to train a machine-learning model in view of the weighted feature.
1. A method comprising:
receiving, by a processor, input data reflecting a feature related to a manufacturing process of a substrate;
generating a characteristic sequence defining a relationship between at least two manufacturing parameters;
determining a relationship between one or more variables related to the feature and the characteristic sequence;
determining a weight based on the determined relationship;
applying the weight to the feature; and
training a machine-learning model in view of the weighted feature.
2. The method of claim 1 , wherein the characteristic sequence comprises a vector.
3. The method of claim 1 , wherein determining the relationship comprises generating a correlation matrix between the one or more variables related to the feature and the characteristic sequence.
4. The method of claim 3 , wherein the correlation matrix is generated using a Pearson correlation method or a Quotient correlation method.
5. The method of claim 1 , wherein the feature relates to an aspect of the manufacturing process that is regulated, monitored, controlled, measured, or associated with one or more sets of sensors or controls.
6. The method of claim 1 , wherein the characteristic sequence is generated by at least one of an algorithm, heuristics data, or a machine-learning model.
7. The method of claim 1 , further comprising:
obtaining, by the processor, sensor data associated with a substrate manufacturing process performed in a process chamber;
inputting the sensor data into the trained machine-learning model; and
obtaining an output value of the trained machine-learning model, the output value being indicative of metrology data.
8. The method of claim 7 , further comprising:
performing a corrective action or updating a process recipe based on the output value.
9. An electronic device manufacturing system, comprising:
a memory device; and
a processing device, operatively coupled to the memory device, to perform operations comprising:
receiving input data reflecting a feature related to a manufacturing process of a substrate;
generating a characteristic sequence defining a relationship between at least two manufacturing parameters;
determining a relationship between one or more variables related to the feature and the characteristic sequence;
determining a weight based on the determined relationship;
applying the weight to the feature; and
training a machine-learning model in view of the weighted feature.
10. The electronic device manufacturing system of claim 9 , wherein the characteristic sequence comprises a vector.
11. The electronic device manufacturing system of claim 9 , wherein determining the relationship comprises generating a correlation matrix between the one or more variables related to the feature and the characteristic sequence.
12. The electronic device manufacturing system of claim 11 , wherein the correlation matrix is generated using a Pearson correlation method or a Quotient correlation method.
13. The electronic device manufacturing system of claim 9 , wherein the feature relates to an aspect of the manufacturing process that is regulated, monitored, controlled, measured, or associated with one or more sets of sensors or controls.
14. The electronic device manufacturing system of claim 9 , wherein the characteristic sequence is generated by at least one of an algorithm, heuristics data, or a machine-learning model.
15. The electronic device manufacturing system of claim 9 , wherein the processing device is further to perform operations comprising:
Obtaining sensor data associated with a substrate manufacturing process performed in a process chamber;
inputting the sensor data into the trained machine-learning model; and
obtaining an output value of the trained machine-learning model, the output value being indicative of metrology data.
16. The electronic device manufacturing system of claim 15 , wherein the processing device is further to perform operations comprising:
performing a corrective action or updating a process recipe based on the output value.
17. A non-transitory computer-readable storage medium comprising instructions that, when executed by a processing device operatively coupled to a memory, performs operations comprising:
receiving input data reflecting a feature related to a manufacturing process of a substrate;
generating a characteristic sequence defining a relationship between at least two manufacturing parameters;
determining a relationship between one or more variables related to the feature and the characteristic sequence;
determining a weight based on the determined relationship;
applying the weight to the feature; and
training a machine-learning model in view of the weighted feature.
18. The non-transitory computer-readable storage medium of claim 17 , wherein the characteristic sequence comprises a vector.
19. The non-transitory computer-readable storage medium of claim 17 , wherein determining the relationship comprises generating a correlation matrix between the one or more variables related to the feature and the characteristic sequence.
20. The non-transitory computer-readable storage medium of claim 19 , wherein the processing device is further to perform operations comprising:
obtaining sensor data associated with a substrate manufacturing process performed in a process chamber;
inputting the sensor data into the trained machine-learning model; and
obtaining an output value of the trained machine-learning model, the output value being indicative of metrology data.