UNPKG

predict-vbb-delays

Version:

Predict delays of Berlin public transport vehicles.

38 lines (24 loc) 1.94 kB
# predict-vbb-delays **Predict delays of Berlin public transport vehicles.** Work in progress, made during [Hamburg Mobility Hackathon 2017](http://mobility-hackathon.de). We generate predictions in multiple stages: 1. **Record departures** with realtime delays, e.g. using [record-vbb-delays](https://github.com/derhuerst/record-vbb-delays#record-vbb-delays). 2. **Transform the data into a training set**, which is just a list of individual data points. Each data point consists of the input vector (one value per feature to be recognized) and a desired output value. `node index.js` will generate an [ndjson](http://ndjson.org) training set. 3. Split the raw training set into **two disjunct sets, one for training, one for auditing** the predictions made later. `node split-into-sets.js` will do this. 4. **Train a neural network**. *More details to be added by @juliuste.* 5. **Generate predictions by applying the trained network** to current realtime departures or future scheduled departures. *Not done yet.* [![npm version](https://img.shields.io/npm/v/predict-vbb-delays.svg)](https://www.npmjs.com/package/predict-vbb-delays) [![build status](https://img.shields.io/travis/derhuerst/predict-vbb-delays.svg)](https://travis-ci.org/derhuerst/predict-vbb-delays) ![ISC-licensed](https://img.shields.io/github/license/derhuerst/predict-vbb-delays.svg) [![chat on gitter](https://badges.gitter.im/derhuerst.svg)](https://gitter.im/derhuerst) ## Installing ```shell git clone https://github.com/derhuerst/predict-vbb-delays.git cd predict-vbb-delays npm i ``` You should now be able to run the scripts in this repo. ## Usage ```js todo ``` ## Contributing If you have a question or have difficulties using `predict-vbb-delays`, please double-check your code and setup first. If you think you have found a bug or want to propose a feature, refer to [the issues page](https://github.com/derhuerst/predict-vbb-delays/issues).