Multi-dimensional safety model

A method of using a machine learning model to determine a multi-dimensional safety indicator, comprising:

• receiving, at an analysis system and from a user device, an indication of a property;
• mapping, by the analysis system, the indication of the property to a plurality of possible locations of a plurality of possible schools associated with the property;
• outputting, from the analysis system and to the user device, an indication of the plurality of possible locations;
• receiving, at the analysis system and from the user device, an indication of a selected location from the plurality of possible locations;
• receiving, from a data source, a set of feedback associated with the selected location;
• verifying information associated with the set of feedback based on:
• transmitting, to a database, a query indicating a set of sources associated with the set of feedback;
• and
• receiving, from the database, a response indicating that the set of sources are verified;
• providing a representation of the selected location to the machine learning model in order to receive the multi-dimensional safety indicator, wherein the multi-dimensional safety indicator comprises a plurality of scores associated with a respective plurality of dimensions,
• wherein the machine learning model is trained using a cross-validation technique comprising:
• splitting a training set into a number of groups including a hold-out group and a training group;
• and
• training the machine learning model on the training group and then testing the machine learning model on the hold-out group to generate a cross-validation score;
• and
• outputting, from the analysis system and to the user device, a data structure encoding the multi-dimensional safety indicator.

A non-transitory computer-readable medium storing a set of instructions for receiving a multi-dimensional safety indicator determined by a machine learning model, the set of instructions comprising:

• one or more instructions that, when executed by one or more processors of a device, cause the device to:
• transmit, to an analysis system, an indication of a property;
• receive, from the analysis system and in response to the indication of the property, an indication of a plurality of possible locations of a plurality of possible schools associated with the property;
• transmit, to the analysis system, an indication of a selected location from the plurality of possible locations;
• and
• receive, from the analysis system and in response to the indication of the selected location, a data structure encoding the multi-dimensional safety indicator associated with the selected location, wherein the multi-dimensional safety indicator comprises a plurality of scores that were determined by the machine learning model and are associated with a respective plurality of dimensions,
• wherein information associated with a set of feedback associated with the selected location was verified based on:
• transmitting, to a database, a query indicating a set of sources associated with the set of feedback;
• and
• receiving, from the database, a response indicating that the set of sources are verified;
• wherein a representation of the selected location is provided to the machine learning model in order to receive the data structure encoding the multi-dimensional safety indicator,
• wherein the machine learning model is trained using a cross-validation technique comprising:
• splitting a training set into a number of groups including a hold-out group and a training group;
• and
• training the machine learning model on the training group and then testing the machine learning model on the hold-out group to generate a cross-validation score.