Systems and methods for dynamically selecting transportation options based on transportation network conditions

A system comprising:

• an identification module, stored in memory, that identifies a transportation requestor device that has initiated a session with a dynamic transportation matching system prior to making a request for transportation via a dynamic transportation network;
• a determination module, stored in memory, that:
• trains a neural network in a first stage using a first set of training data representing transportation options that are available through the dynamic transportation network;
• trains the neural network in a second stage using a second set of training data representing potential impacts to the dynamic transportation network of different selected transportation options, such that the neural network is trained to dynamically estimate an impact of a plurality of transportation options on the dynamic transportation network;
• and
• determines a current condition of the dynamic transportation network at least in part by examining, by the dynamic transportation matching system, a current status of at least one transportation provider within the dynamic transportation network that affects the current condition of the dynamic transportation network;
• a selection module, stored in memory, that identifies, via the trained neural network and based at least in part on the estimated impact of the available transportation options on the current condition of the dynamic transportation network, which of the available transportation options is estimated to reduce impact to the dynamic transportation network;
• a sending module, stored in memory, that sends, to the transportation requestor device for display on the transportation requestor device, those transportation options identified by the neural network to complete the transportation request;
• and
• at least one physical processor that executes the identification module, the determination module, the selection module, and the sending module.

A computer-readable medium comprising:

• computer-readable instructions that, when executed by at least one processor of a computing device, cause the computing device to:
• identify a transportation requestor device that has initiated a session with a dynamic transportation matching system prior to making a request for transportation via a dynamic transportation network;
• train a neural network in a first stage using a first set of training data representing transportation options that are available through the dynamic transportation network;
• training the neural network in a second stage using a second set of training data representing potential impacts to the dynamic transportation network of different selected transportation options, such that the neural network is trained to dynamically estimate an impact of a plurality of available transportation options on the dynamic transportation network;
• determine a current condition of the dynamic transportation network at least in part by examining, by the dynamic transportation matching system, a current status of at least one transportation provider within the dynamic transportation network that affects the current condition of the dynamic transportation network;
• identify, via the trained neural network and based at least in part on the estimated impact of the available transportation options on the current condition of the dynamic transportation network, which of the available transportation options is estimated to reduce impact to the dynamic transportation network;
• and
• send, to the transportation requestor device for display on the transportation requestor device, those transportation options identified by the neural network to complete the transportation request.