Method and system of automatedly generating insurance coverage recommendations using predicted weather

A computing system for automatically generating insurance coverage recommendations, the computing system comprising:

• one or more processors;
• one or more memories storing instructions that, when executed by the one or more processors, cause the computing system to perform a process comprising:
• retrieving user data evincing a user intent for a planned relocation of a user;
• determining a prediction of weather for a location and duration of the planned relocation of the user;
• determining a type of user risk indicating a level of risk taking or aversion of the user, derived from prior observed behavior of the user;
• generating a user activity profile for the user comprising at least i) a first indication of the predicted weather for the planned relocation, ii) a second indication of the type of user risk, and iii) a third indication of a type of the planned relocation, wherein the first and second indications are each paired with a corresponding degree of severity;
• converting the user activity profile into input for the machine learning model by entering, into a vector, at least i) the first indication of the predicted weather for the planned relocation and ii) the second indication of the type of user risk,
• wherein the machine learning model is trained, via supervised learning and in a training phase, to predict one or more needed insurance coverage levels for the planned relocation, based on training data comprising A) insurance incident claims records defining training user activity profiles, each training user activity profile including at least weather data at the time of the corresponding incident paired with B) needed insurance coverage levels to cover losses for the corresponding incident,
• wherein the machine learning model establishes a probability distribution resulting from analysis of the training data, the probability distribution being a likelihood of an input planned relocation matching a particular needed insurance coverage level, and
• wherein the particular needed insurance coverage level, having a highest probability according to the probability distribution, is compared to a desired output for a paired insurance coverage level and, based on the comparison, the machine learning model is modified in the training phase;
• customizing the trained machine learning model and applying the customized machine learning model to the input by:
• applying customizations, defined by X) the corresponding degree of severity for the first indication of the predicted weather for the planned activity, Y) the corresponding degree of severity for the second indication of the type of user risk, and Z) the third indication of the type of the planned relocation, to the trained machine learning model causing the machine learning model to respond to the input specifically for similar relocations, similar weather severity, and similar user risk severity;
• and
• applying the input to the customized machine learning model;
• in response to the applying, obtaining the one or more needed insurance coverage levels for the planned relocation;
• and
• generating a user interface including one or more insurance coverage recommendations respectively corresponding to the one or more needed insurance coverage levels for the planned relocation.

A machine-readable storage medium having machine-executable instructions stored thereon that, when executed by one or more processors, cause the one or more processors to perform a method for automatically generating insurance coverage recommendations, the method comprising:

• retrieving user data evincing a user intent for a planned relocation of a user;
• determining a prediction of weather for a location and duration of the planned relocation of the user;
• determining a type of user risk indicating a level of risk taking or aversion of the user, derived from prior observed behavior of the user;
• generating a user activity profile for the user comprising at least i) a first indication of the predicted weather for the planned relocation, ii) a second indication of the type of user risk, and iii) a third indication of a type of the planned relocation, wherein the first and second indications are each paired with a corresponding degree of severity;
• converting the user activity profile into input for a machine learning model by entering, into a vector, at least i) the first indication of the predicted weather for the planned relocation and ii) the second indication of the type of user risk,
• wherein the machine learning model is trained, via supervised learning and in a training phase, to predict one or more needed insurance coverage levels for the planned relocation, based on training data comprising A) insurance incident claims records defining training user activity profiles, each training user activity profile including at least weather data at the time of the corresponding incident paired with B) needed insurance coverage levels to cover losses for the corresponding incident,
• wherein the machine learning model establishes a probability distribution resulting from analysis of the training data, the probability distribution being a likelihood of an input planned relocation matching a particular needed insurance coverage level, and
• wherein the particular needed insurance coverage level, having a highest probability according to the probability distribution, is compared to a desired output for a paired insurance coverage level and, based on the comparison, the machine learning model is modified in the training phase;
• customizing the trained machine learning model and applying the customized machine learning model to the input by:
• applying customizations, defined by X) the corresponding degree of severity for the first indication of the predicted weather for the planned relocation, Y) the corresponding degree of severity for the second indication of the type of user risk, and Z) the third indication of the type of the planned relocation, to the trained machine learning model causing the machine learning model to respond to the input specifically for similar relocations, similar weather severity, and similar user risk severity;
• and
• applying the input to the customized machine learning model;
• in response to the applying, obtaining the one or more needed insurance coverage levels for the planned relocation;
• and
• generating a user interface including one or more insurance coverage recommendations respectively corresponding to the one or more needed insurance coverage levels for the planned relocation.