System and method for model-agnostic meta-learner for noisy data with label errors

1. A method implemented on at least one machine including at least one processor, memory, and communication platform capable of connecting to a network for machine learning, the method comprising:

performing machine learning based on training data via a dual loop learning process having a first loop for data decoding learning and a second loop for label decoding learning for correcting labeling errors in training data provided by a source, wherein

in the first loop, first parameters associated with the data decoding are updated to generate updated first parameters based on a first label estimated via the data decoding using the first parameters and a second label predicted via the label decoding using second parameters, wherein the second label is a predicted ground truth label, and

in the second loop, the second parameters associated with the label decoding are updated to generate updated second parameters based on a third label obtained via the data decoding using the updated first parameters and a correctly labeled ground truth label provided by a target, wherein the updated second parameters are used to predict the second label in the first loop based on the training data including the labeling errors.

2. The method of claim 1 , wherein the first loop comprises:

receiving first input data from the source;

classifying, based on the first parameters, the first input data to have the first label;

receiving the second label predicted by the label decoding based on the input data and second parameters;

determining a source loss based on the first label and the second label; and

generating the updated first parameters based on the source loss.

3. The method of claim 2 , wherein the source provides training data with inconsistent, incorrect, and missing labels.

4. The method of claim 2 , wherein the first parameters associated with the data decoding include encoder parameters and decoder parameters.

5. The method of claim 4 , wherein the step of classifying comprises:

encoding, based on the encoder parameters, the first input data to generate a feature vector; and

decoding, based on the decoder parameters, the first input data in accordance with the encoded feature vector to generate the first label.

6. The method of claim 1 , wherein the second loop comprises:

receiving second input data from the target, wherein the second input is associated with the correctly labeled ground truth label;

classifying the second input data to have the third label based on the updated first parameters;

determining a target loss based on the third label and the correctly labeled ground truth label; and

updating the second parameters based on the target loss.

7. The method of claim 6 , wherein the target provides training data with the correctly labeled ground truth label.

8. Machine readable and non-transitory medium having information recorded thereon for machine learning, wherein the information, once read by the machine, causes the machine to perform:

performing machine learning based on training data via a dual loop learning process having a first loop for data decoding learning and a second loop for label decoding learning for correcting labeling errors in training data provided by a source, wherein

in the first loop, first parameters associated with the data decoding are updated to generate updated first parameters based on a first label estimated via the data decoding using the first parameters and a second label predicted via the label decoding using second parameters, wherein the second label is a predicted ground truth label, and

in the second loop, the second parameters associated with the label decoding are updated to generate updated second parameters based on a third label obtained via the data decoding using the updated first parameters and a correctly labeled ground truth label provided by a target, wherein the updated second parameters are used to predict the second label in the first loop based on the training data including the labeling errors.

9. The medium of claim 8 , wherein steps performed in the first loop comprises:

receiving first input data from the source;

classifying, based on the first parameters, the first input data to have the first label;

receiving the second label predicted by the label decoding based on the input data and second parameters;

determining a source loss based on the first label and the second label; and

generating the updated first parameters based on the source loss.

10. The medium of claim 9 , wherein the source provides training data with inconsistent, incorrect, and missing labels.

11. The medium of claim 9 , wherein the first parameters associated with the data decoding include encoder parameters and decoder parameters.

12. The medium of claim 11 , wherein the step of classifying comprises:

encoding, based on the encoder parameters, the first input data to generate a feature vector; and

decoding, based on the decoder parameters, the first input data in accordance with the encoded feature vector to generate the first label.

13. The medium of claim 8 , wherein steps performed in the second loop comprises:

receiving second input data from the target, wherein the second input is associated with the correctly labeled ground truth label;

classifying the second input data to have the third label based on the updated first parameters;

determining a target loss based on the third label and the correctly labeled ground truth label; and

updating the second parameters based on the target loss.

14. The medium of claim 13 , wherein the target provides training data with the correctly labeled ground truth label.

15. A system for machine learning, comprising:

the system comprising:

memory storing computer program instructions; and

one or more processors that, in response to executing the computer program instructions, effectuate operations comprising:

performing machine learning based on training data via a dual loop learning process having a first loop for data decoding learning and a second loop for label decoding learning for correcting labeling errors in training data provided by a source wherein

in the first loop, first parameters associated with the data decoding are updated to generate updated first parameters based on a first label classified by the data decoding using the first parameters and a second label predicted by the label decoding using second parameters, wherein the second label is a predicted ground truth label, and

in the second loop, the second parameters associated with the label decoding are updated to generate updated second parameters based on a third label classified by the data decoding using the updated first parameters and a correctly labeled ground truth label provided by a target, wherein the updated second parameters are used to predict the second label in the first loop based on the training data including the labeling errors.

16. The system of claim 15 , wherein the first parameters include encoder parameters used by an encoder to encode the training data and decoder parameters used by a decoder to classify the encoded training data.

17. The system of claim 16 , wherein the first loop comprises:

receiving a first input data from the source;

encoding the first input data to generated encoded first input data based on the encoder parameters; and

classifying, based on the first parameters, the first input data to have the first label.

18. The system of claim 17 , wherein the operations further comprise:

determining a source loss based on the first label and the second label; and

updating the first parameters based on the source loss to generate the updated first parameters.

19. The system of claim 17 , wherein the operations comprise:

in the first loop, predicting the second label for the first input data based on the second parameters;

in the second loop, retrieving the correctly labeled ground truth label for second input data from the target;

in the second loop, computing a target loss based on the correctly labeled ground truth label and the third label generated by the decoder based on the second input data in accordance with the updated first parameters; and

in the second loop, updating the second parameters based on the target loss to generate the updated second parameters.

20. The system of claim 19 , wherein

the source provides training data with inconsistent, incorrect, and missing labels; and

the target provides training data with the correctly labeled ground truth label.