Configurable predictive models for account scoring and signal synchronization

One or more non-transitory computer-readable media storing instructions which, when executed by one or more hardware processors cause:

• generating, using a model, a first set of scores for a plurality of accounts, wherein a score in first set of scores for a respective account characterizes a priority of interacting with the respective account relative to other accounts in the plurality of accounts, wherein the model defines a set of criteria for scoring accounts, wherein each criterion in the set of criteria is associated with one or more feature values and a weight;
• determining a distribution of the plurality of accounts based at least in part on the first set of scores and an outcome of a targeted action;
• tuning, by a machine learning process, the model based at least in part on the distribution of the plurality of accounts relative to the outcome of the targeted action, wherein tuning the set of criteria comprises iteratively adjusting at least one weight associated with at least one feature value until the distribution of the plurality of accounts satisfies at least one threshold, wherein the threshold is determined based at least in part on positions of data points in the distribution relative to positions of data points in a random distribution;
• generating, using the tuned model, a second set of scores for the plurality of accounts;
• and
• performing at least one action using a set of one or more computing resources with respect to a subset of the plurality of accounts based at least in part on the second set of scores, wherein the at least one action is prioritized by the set of one or more computing resources based at least in part on the second set of scores.

A system comprising:

• one or more hardware processors;
• one or more non-transitory computer-readable media storing instructions which, when executed by the one or more hardware processors cause:
• generating, using a model, a first set of scores for a plurality of accounts, wherein a score in first set of scores for a respective account characterizes a priority of interacting with the respective account relative to other accounts in the plurality of accounts, wherein the model defines a set of criteria for scoring accounts, wherein each criterion in the set of criteria is associated with one or more feature values and a weight;
• determining a distribution of the plurality of accounts based at least in part on the first set of scores and an outcome of a targeted action;
• tuning, by a machine learning process, the model based at least in part on the distribution of the plurality of accounts relative to the outcome of the targeted action, wherein tuning the set of criteria comprises iteratively adjusting at least one weight associated with at least one feature value until the distribution of the plurality of accounts satisfies at least one threshold, wherein the threshold is determined based at least in part on positions of data points in the distribution relative to positions of data points in a random distribution;
• generating, using the tuned model, a second set of scores for the plurality of accounts;
• and
• performing at least one action using a set of one or more computing resources with respect to a subset of the plurality of accounts based at least in part on the second set of scores, wherein the at least one action is prioritized by the set of one or more computing resources based at least in part on the second set of scores.