location-classifier v1.0.2

Location Classifier - AI That Classifies Locations as Accessible or Inaccessible

We highly suggest you check out our GitHub Repository. It will have more information about our project.

Sponsored by XI Media LLC, this project is goal one of four for my groups senior design project. The purpose of this project is to create a system using artificial intelligence that can randomly select an accessible location in a region on a map. The underlying problem is a binary classification problem where the input is a location (latitude, longitude) and the output is binary: it is either accessible or inaccessible. To solve this problem, our team designed a system that builds upon the TensorFlowJS library. Our system will take an input location, transform the location into interpretable data, vectorize the interpretable data, feed the input vector into a TensorFlow model, and apply a threshold to determine the binary result.

This system is two-fold: it acts as a NodeJS package for developers who wish to use artificial intelligence to classify locations as accessible or inaccessible, and it acts as a toolkit for future development teams by providing all of the necessary routines to create models, introduce new datasets, and analyze flaws in current iterations.

Example Usage

let API = require('location-classifier');
let lc = new API();

...

let is_accessible = await lc.pipeline(latitude, longitude);

To learn more about the parameters and methods, please visit our GitHub Repository.