CPC G06Q 10/08355 (2013.01) [G06N 20/00 (2019.01)] | 16 Claims |
1. A method comprising:
receiving first data associated with one or more first deliveries performed by a first delivery driver at a first time;
training a computing model including a decision tree using the first data;
receiving, at a second time, a delivery itinerary including one or more second deliveries to be performed by the first delivery driver or a second delivery driver;
receiving second data associated with the one or more second deliveries, the second data including delivery driver data, customer data, address data, delivery package data, seller data, and environmental data, wherein the second data comprises a first data table that includes a data row including a first identifier and a second data row also including the first identifier;
performing a first data transformation on the first data table to generate a second data table including a single data row including the first identifier;
performing a second data transformation on the second data table such that a single historical delivery occupies a single row of the second data table, wherein the second data transformation involves removing a first column from the first data table and aggregating a second column and a third column including non-numerical values into a different column type including numerical values;
determining, in real-time, using the computing model, and based on the delivery itinerary and the second data, a probability that the one or more second deliveries will result in a delivery defect, wherein the delivery defect includes at least one of: a package that is delivered but not received by a customer, a package that is delivered to an incorrect address, or a failure to satisfy a customer delivery request;
determining, based on the probability being greater than a first threshold value, a first modification to a first delivery parameter for the one or more second deliveries, wherein the first modification involves adding a first limitation to the one or more second deliveries; and
presenting the first limitation through a user interface of a mobile device application used by the first delivery driver or the second delivery driver to perform the one or more second deliveries.
|