System and method for facilitating prediction model training

Abstract

In certain embodiments, training data may be generated for training a prediction model. Training data including first datasets may be obtained, where the first datasets include a plurality of feature types. A determination, via a relevancy model, based on the training data, of whether a feature type satisfies a first condition may be made. If the first condition is satisfied, one or more second datasets may be obtained to update the training data, where the second datasets include the plurality of feature types. A determination, via the relevancy model, based on the updated training data, may be made as to whether the feature type satisfies a second condition. The first and second conditions may relate to whether the feature type has a threshold amount of influence on the prediction model. If the second condition is satisfied, the updated training data may be provided to the prediction model.

What is claimed is: 
     
         1 . A non-transitory computer-readable medium storing computer program instructions that, when executed by one or more processors, effectuate operations comprising:
 obtaining first training data comprising first datasets, wherein the first datasets comprise a plurality of feature types;   determining, via a relevancy model, based on the first training data, whether a feature type of the plurality of feature types satisfies a first condition, the first condition being satisfied comprising a relevancy score of the feature type being equal to or greater than a threshold relevancy score;   responsive to determining that the feature type satisfies the first condition, obtaining one or more second datasets to update the first training data to obtain second training data, wherein the second training data comprises the one or more second datasets, wherein the one or more second datasets comprise the plurality of feature types;   determining, via the relevancy model, based on the second training data, whether the feature type satisfies a second condition, the second condition being satisfied comprising an updated relevancy score of the feature type being less than the threshold relevancy score; and   responsive to determining that the feature type satisfies the second condition, causing a machine learning model to be trained with the second training data.   
     
     
         2 . The non-transitory computer-readable medium of  claim 1 , wherein the second training data does not comprise one or more other ones of the first datasets. 
     
     
         3 . The non-transitory computer-readable medium of  claim 1 , wherein the operations further comprise:
 identifying a set of feature types that are to be prevented from having a threshold amount of influence on the machine learning model, wherein the set of feature types comprises the feature type.   
     
     
         4 . The non-transitory computer-readable medium of  claim 3 , wherein the operations further comprise:
 determining, prior to the second training data being used to train the machine learning model, that each feature type of the set of feature types satisfies the second condition.   
     
     
         5 . The non-transitory computer-readable medium of  claim 1 , wherein:
 obtaining the first training data comprises randomly selecting the first datasets from one or more data corpora; and   obtaining the one or more second datasets comprises randomly selecting the one or more second datasets from the one or more data corpora.   
     
     
         6 . The non-transitory computer-readable medium of  claim 1 , wherein:
 the machine learning model comprises a neural network; and   the relevancy model comprises a principle component analysis (PCA) model.   
     
     
         7 . The non-transitory computer-readable medium of  claim 1 , wherein the operations further comprise:
 determining, via the relevancy model, based on the second training data, whether an additional feature type of the plurality of feature types satisfies the first condition;   responsive to determining that the additional feature type satisfies the first condition, obtaining one or more third datasets to update the second training data to obtain third training data; and   determining, via the relevancy model, based on the third training data, whether the additional feature type satisfies the second condition, wherein whether the feature type satisfies the second condition is determined responsive to the additional feature type being determined to satisfies the second condition.   
     
     
         8 . A system, comprising:
 memory storing computer program instructions; and   one or more processors that, in response to executing the computer program instructions, effectuate operations comprising:
 obtaining first training data comprising first datasets, wherein the first datasets comprise a plurality of feature types; 
 determining, via a relevancy model, based on the first training data, whether a feature type of the plurality of feature types satisfies a first condition, the first condition being satisfied comprising a relevancy score of the feature type being equal to or greater than a threshold relevancy score; 
 responsive to determining that the feature type satisfies the first condition, obtaining one or more second datasets to update the first training data to obtain second training data, wherein the second training data comprises the one or more second datasets, wherein the one or more second datasets comprise the plurality of feature types; 
 determining, via the relevancy model, based on the second training data, whether the feature type satisfies a second condition, the second condition being satisfied comprising an updated relevancy score of the feature type being less than the threshold relevancy score; and 
   responsive to determining that the feature type satisfies the second condition, causing a machine learning model to be trained with the second training data.   
     
     
         9 . The system of  claim 8 , wherein the second training data does not comprise one or more other ones of the first datasets. 
     
     
         10 . The system of  claim 8 , wherein the operations further comprise:
 identifying a set of feature types that are to be prevented from having a threshold amount of influence on the machine learning model, wherein the set of feature types comprises the feature type.   
     
     
         11 . The system of  claim 10 , wherein the operations further comprise:
 determining, prior to the second training data being used to train the machine learning model, that each feature type of the set of feature types satisfies the second condition.   
     
     
         12 . The system of  claim 8 , wherein:
 obtaining the first training data comprises randomly selecting the first datasets from one or more data corpora; and   obtaining the one or more second datasets comprises randomly selecting the one or more second datasets from the one or more data corpora.   
     
     
         13 . The system of  claim 8 , wherein:
 the machine learning model comprises a neural network; and   the relevancy model comprises a principle component analysis (PCA) model.   
     
     
         14 . The system of  claim 8 , wherein the operations further comprise:
 determining, via the relevancy model, based on the second training data, whether an additional feature type of the plurality of feature types satisfies the first condition;   responsive to determining that the additional feature type satisfies the first condition, obtaining one or more third datasets to update the second training data to obtain third training data; and   determining, via the relevancy model, based on the third training data, whether the additional feature type satisfies the second condition, wherein whether the feature type satisfies the second condition is determined responsive to the additional feature type being determined to satisfies the second condition.   
     
     
         15 . A method implemented by one or more processors configured to execute computer program instructions, the method comprising:
 obtaining first training data comprising first datasets, wherein the first datasets comprise a plurality of feature types;   determining, via a relevancy model, based on the first training data, whether a feature type of the plurality of feature types satisfies a first condition, the first condition being satisfied comprising a relevancy score of the feature type being equal to or greater than a threshold relevancy score;   responsive to determining that the feature type satisfies the first condition, obtaining one or more second datasets to update the first training data to obtain second training data, wherein the second training data comprises the one or more second datasets, wherein the one or more second datasets comprise the plurality of feature types;   determining, via the relevancy model, based on the second training data, whether the feature type satisfies a second condition, the second condition being satisfied comprising an updated relevancy score of the feature type being less than the threshold relevancy score; and   responsive to determining that the feature type satisfies the second condition, causing a machine learning model to be trained with the second training data.   
     
     
         16 . The method of  claim 15 , further comprising:
 identifying a set of feature types that are to be prevented from having a threshold amount of influence on the machine learning model, wherein the set of feature types comprises the feature type.   
     
     
         17 . The method of  claim 16 , further comprising:
 determining, prior to the second training data being used to train the machine learning model, that each feature type of the set of feature types satisfies the second condition.   
     
     
         18 . The method of  claim 15 , wherein:
 obtaining the first training data comprises randomly selecting the first datasets from one or more data corpora; and   obtaining the one or more second datasets comprises randomly selecting the one or more second datasets from the one or more data corpora.   
     
     
         19 . The method of  claim 15 , wherein:
 the second training data does not comprise one or more other ones of the first datasets;   the machine learning model comprises a neural network; and   the relevancy model comprises a principle component analysis (PCA) model.   
     
     
         20 . The method of  claim 15 , further comprising:
 determining, via the relevancy model, based on the second training data, whether an additional feature type of the plurality of feature types satisfies the first condition;   responsive to determining that the additional feature type satisfies the first condition, obtaining one or more third datasets to update the second training data to obtain third training data; and   determining, via the relevancy model, based on the third training data, whether the additional feature type satisfies the second condition, wherein whether the feature type satisfies the second condition is determined responsive to the additional feature type being determined to satisfies the second condition.