System and method for processing optimization through similarity analysis

Abstract

This disclosure provides a system and method for process optimization through similarity analysis, which selects at least one historical process data that is closest to the target process from historical dataset, adjusts parameters based on the historical process data to obtain the target process, and then verifies the accuracy of the target process to ensure the quality of products produced using the target process. Not only can it reduce the cost of testing and production risks, but also systematically adjust the weight of parameters to produce a suitable process for subsequent production, thereby shortening development time and improving yield.

Claims

1. A process optimization system based on similarity analysis, comprising: Storage, storing multiple modules; and a processor, coupled to the storage and configured to: acquire multiple data corresponding to multiple specifications; The preprocessing module in these modules is executed to adjust the weights corresponding to multiple factors in these specifications; the encoding module in these modules is executed to encode these factors; the comparison module in these modules is executed to calculate multiple similarities based on the target process and the encoded factors; the analysis module in these modules is executed to obtain target data based on the similarities and calculate multiple target process factors corresponding to the target process based on the target data; an optimized target process is established based on these target process factors; and the prediction module in these modules is executed to predict the accuracy of the optimized target process.

2. The process optimization system via similarity analysis as described in claim 1, wherein the weights correspond to the use case of the target process.

3. The process optimization system via similarity analysis as described in claim 1, wherein the factors include size, material, and structure.

4. The process optimization system via similarity analysis as described in claim 1, wherein the processor is further configured to: encode the factors using label encoding, one-hot encoding, or mean encoding.

5. The process optimization system via similarity analysis as described in claim 1, wherein the processor is further configured to: calculate multiple similarities based on multiple target process factors of the target process and multiple data factors of each of the data.

6. The process optimization system via similarity analysis as described in claim 5, wherein the processor is further configured to: calculate the similarity for factors belonging to numerical variables using Euclidean distance; and calculate the similarity for factors belonging to categorical variables using Jaccard similarity or cosine similarity.

7. The process optimization system via similarity analysis as described in claim 6, wherein the processor is further configured to: establish the target data based on first data among the data, wherein the first data has a first overall similarity to the target process, wherein the second data among the data has a second overall similarity to the target process, wherein the first overall similarity is greater than the second overall similarity.

8. The process optimization system via similarity analysis as described in claim 6, wherein the processor is further configured to: establish the target data based on a plurality of third data among the data, wherein each of the third data has a third overall similarity to the target process, wherein the third overall similarity is greater than a first threshold.

9. The process optimization system via similarity analysis as described in claim 1, wherein the processor is further configured to: adjust the weights according to the detailed use case of the target process in response to the accuracy being less than a second threshold.

10. A method for process optimization through similarity analysis, executed by a processor, the method comprising: Obtain multiple data points corresponding to multiple specifications; adjust the weights of multiple factors within these specifications; encode these factors; Calculate multiple similarities based on the target process and the encoded factors; obtain target data based on the similarities, and calculate multiple target process factors corresponding to the target process based on the target data; establish an optimized target process based on the target process factors; and predict the accuracy of the optimized target process.

11. The process optimization method via similarity analysis as described in claim 10, wherein the weights correspond to the use case of the target process.

12. The process optimization method via similarity analysis as described in claim 10, wherein the factors include size, material, and structure.

13. The process optimization method using similarity analysis as described in claim 10 further includes: These factors are encoded using label coding, read-hot coding, or mean coding.

14. The process optimization method using similarity analysis as described in claim 10 further includes: Multiple similarities are calculated based on multiple target process factors of the target process and multiple data factors of each of these data.

15. The process optimization method using similarity analysis as described in claim 14 further includes: The similarity is calculated using Euclidean distance for factors belonging to numerical variables; and using Jaccard similarity or cosine similarity for factors belonging to categorical variables.

16. The process optimization method using similarity analysis as described in claim 15 further includes: The target data is established based on the first data in the data, wherein the first data has a first overall similarity with the target process, wherein the second data in the data has a second overall similarity with the target process, and wherein the first overall similarity is greater than the second overall similarity.

17. The process optimization method using similarity analysis as described in claim 15 further includes: The target data is established based on multiple third data from these data, wherein each of these third data has a third overall similarity with the target process, and the third overall similarity is greater than a first threshold.

18. The process optimization method using similarity analysis as described in claim 10 further includes: In response to the accuracy rate being less than the second threshold, these weights are adjusted according to the specific use case of the target process.

Images