Predictive data analysis using value-based predictive inputs

Abstract

There is a need for solutions that perform predictive data analysis using a value-based predictive input. This need can be addressed by, for example, determining, based at least in part on the value-based predictive input, a plurality of predictive component values; for each predictive component value of the plurality of predictive component values: obtaining a quantile regression distribution for the predictive component value; determining, based at least in part on the quantile regression distribution, a non-outlier portion of the quantile regression distribution; generating, for each quantile regression value of the one or more quantile regression values that is associated with the non-outlier portion, a scaled quantile regression value; and determining, based at least in part on each scaled quantile regression value for a quantile regression value associated with a predictive component value of the plurality of predictive component values, an entity opportunity prediction of the one or more entity predictions for the prediction entity.

1 . An apparatus comprising one or more processors and one or more non-transitory computer-readable storage media including instructions that, when executed by the one or more processors, cause the apparatus to:
 generate entity-level prediction data for a prediction entity based at least in part on raw transactional data and one or more entity-level aggregation rules;   reduce a complexity of the entity-level prediction data;   determine, based at least in part on a value-based predictive input associated with the prediction entity, one or more first predictive component values;   select, from among a plurality of prediction engines and based at least in part on the one or more predictive component values, a first subset of prediction engines with highest quantile regression values as a first most predicted value corresponding to the prediction entity; and   determine, based at least in part on the first subset of prediction engines, one or more first entity predictions for the prediction entity.   
     
     
         2 . The apparatus of  claim 1 , wherein the one or more non-transitory computer-readable storage media include instructions that, when executed by the one or more processors, further cause the apparatus to:
 determine, based at least in part on the one or more prediction entities, one or more second predictive component values;   select, from among the plurality of prediction engines and based at least in part on the one or more second predictive component values, a second subset of prediction engines with highest quantile regression values as a second most predicted value corresponding to the prediction entity; and   determine, based at least in part on the second subset of prediction engines, one or more second entity predictions for the prediction entity.   
     
     
         3 . The apparatus of  claim 2 , wherein the one or more second entity predictions represent increased granularity as compared to the one or more first entity predictions. 
     
     
         4 . The apparatus of  claim 1 , wherein the one or more non-transitory computer-readable storage media include instructions that, when executed by the one or more processors, further cause the apparatus to:
 present a prediction report associated with the one or more first entity predictions to a user device.   
     
     
         5 . The apparatus of  claim 2 , wherein the one or more non-transitory computer-readable storage media include instructions that, when executed by the one or more processors, further cause the apparatus to:
 present a prediction report associated with the one or more second entity predictions to a user device.   
     
     
         6 . The apparatus of  claim 1 , wherein the one or more non-transitory computer-readable storage media include instructions that, when executed by the one or more processors, further cause the apparatus to:
 execute a machine learning algorithm trained using gradient descent.   
     
     
         7 . The apparatus of  claim 6 , wherein executing the machine learning algorithm comprises:
 generating aggregated entity-level data for the prediction entity based at least in part on aggregating the entity-level prediction data;   generating, based at least in part on the aggregated entity-level data, the value-based predictive input;   for each predictive component value of the plurality of predictive component values:   obtaining a quantile regression distribution for the predictive component value, wherein the quantile regression distribution indicates a distribution of a corresponding predictive component that is associated with the predictive component value across the plurality of prediction entities via a plurality of quantile regression values,   determining a non-minimum ratio for the quantile regression distribution as a ratio of a non-minimum portion of the quantile regression distribution that falls below or equals a minimum threshold value,   determining a non-outlier ratio for the quantile regression distribution based on a deviation between a full ratio and a product of the non-minimum ratio and an outlier parameter,   determining a non-outlier portion of the quantile regression distribution as a subset of the quantile regression distribution that comprises each segment of the quantile regression distribution whose respective quantile regression value fall below or equals the non-outlier ratio, and   generating, for each quantile regression value of the plurality of quantile regression values that is in the non-outlier portion, respective scaled quantile regression values based on the quantile regression value, the predictive component value, and a quantile regression ratio for the quantile regression value; and   providing the respective scaled quantile regression values to the plurality of prediction engines.   
     
     
         8 . The apparatus of  claim 1 , wherein reducing the complexity of the entity-level prediction data comprises selecting a subset of prediction engines with highest scaled regression values. 
     
     
         9 . The apparatus  claim 1 , wherein the one or more non-transitory computer-readable storage media include instructions that, when executed by the one or more processors, further cause the apparatus to:
 determine an outlier portion for a particular quantile regression distribution associated with a particular predictive component value of the plurality of predictive component values, by:   determining a minimal ratio of the plurality of quantile regression values that exceed a minimal prediction threshold;   determining the outlier parameter for the quantile regression distribution; and   determining the outlier portion based at least in part on the minimal ratio and the outlier parameter.   
     
     
         10 . The apparatus  claim 1 , wherein the one or more non-transitory computer-readable storage media include instructions that, when executed by the one or more processors, further cause the apparatus to:
 determine, based at least in part on the value-based predictive input for the prediction entity, an entity closure prediction of the one or more action-based predictive outputs.   
     
     
         11 . A computer-implemented method, comprising:
 generating, by one or more processors, entity-level prediction data for a prediction entity based at least in part on raw transactional data and one or more entity-level aggregation rules;   reducing, by the one or more processors, a complexity of the entity-level prediction data;   determining, by the one or more processors and based at least in part on a value-based predictive input associated with the prediction entity, one or more first predictive component values;   selecting, by the one or more processors, from among a plurality of prediction engines and based at least in part on the one or more predictive component values, a first subset of prediction engines with highest quantile regression values as a first most predicted value corresponding to the prediction entity; and   determining, by the one or more processors and based at least in part on the first subset of prediction engines, one or more first entity predictions for the prediction entity.   
     
     
         12 . The computer-implemented method of  claim 11 , further comprising:
 determining, based at least in part on the one or more prediction entities, one or more second predictive component values;   selecting, from among the plurality of prediction engines and based at least in part on the one or more second predictive component values, a second subset of prediction engines with highest quantile regression values as a second most predicted value corresponding to the prediction entity; and   determining, based at least in part on the second subset of prediction engines, one or more second entity predictions for the prediction entity.   
     
     
         13 . The computer-implemented method of  claim 12 , wherein the one or more second entity predictions represent increased granularity as compared to the one or more first entity predictions. 
     
     
         14 . The computer-implemented method of  claim 11 , further comprising:
 presenting a prediction report associated with the one or more first entity predictions to a user device.   
     
     
         15 . The computer-implemented method of  claim 12 , further comprising:
 presenting a prediction report associated with the one or more second entity predictions to a user device.   
     
     
         16 . The computer-implemented method of  claim 11 , further comprising:
 executing a machine learning algorithm trained using gradient descent.   
     
     
         17 . The computer-implemented method of  claim 16 , wherein executing the machine learning algorithm comprises:
 generating aggregated entity-level data for the prediction entity based at least in part on aggregating the entity-level prediction data;   generating, based at least in part on the aggregated entity-level data, the value-based predictive input;   for each predictive component value of the plurality of predictive component values:   obtaining a quantile regression distribution for the predictive component value, wherein the quantile regression distribution indicates a distribution of a corresponding predictive component that is associated with the predictive component value across the plurality of prediction entities via a plurality of quantile regression values,   determining a non-minimum ratio for the quantile regression distribution as a ratio of a non-minimum portion of the quantile regression distribution that falls below or equals a minimum threshold value,   determining a non-outlier ratio for the quantile regression distribution based on a deviation between a full ratio and a product of the non-minimum ratio and an outlier parameter,   determining a non-outlier portion of the quantile regression distribution as a subset of the quantile regression distribution that comprises each segment of the quantile regression distribution whose respective quantile regression value fall below or equals the non-outlier ratio, and   generating, for each quantile regression value of the plurality of quantile regression values that is in the non-outlier portion, respective scaled quantile regression values based on the quantile regression value, the predictive component value, and a quantile regression ratio for the quantile regression value; and   providing the respective scaled quantile regression values to the plurality of prediction engines.   
     
     
         18 . The computer-implemented method of  claim 11 , wherein reducing the complexity of the entity-level prediction data comprises selecting a subset of prediction engines with highest scaled regression values. 
     
     
         19 . The computer-implemented method of  claim 11 , further comprising:
 determining an outlier portion for a particular quantile regression distribution associated with a particular predictive component value of the plurality of predictive component values, by:   determining a minimal ratio of the plurality of quantile regression values that exceed a minimal prediction threshold;   determining the outlier parameter for the quantile regression distribution; and   determining the outlier portion based at least in part on the minimal ratio and the outlier parameter.   
     
     
         20 . The computer-implemented method of  claim 11 , further comprising:
 determining, based at least in part on the value-based predictive input for the prediction entity, an entity closure prediction of the one or more action-based predictive outputs.