https://github.com/grasia/knnp
Time Series Forecasting using K-Nearest Neighbors Algorithm (Parallel approach)
https://github.com/grasia/knnp
knearest-neighbor-algorithm parallel time-series-forecasting
Last synced: 18 days ago
JSON representation
Time Series Forecasting using K-Nearest Neighbors Algorithm (Parallel approach)
- Host: GitHub
- URL: https://github.com/grasia/knnp
- Owner: Grasia
- License: agpl-3.0
- Created: 2020-01-08T12:25:43.000Z (about 6 years ago)
- Default Branch: package-release
- Last Pushed: 2020-01-17T09:58:38.000Z (about 6 years ago)
- Last Synced: 2025-12-09T21:18:51.622Z (3 months ago)
- Topics: knearest-neighbor-algorithm, parallel, time-series-forecasting
- Language: R
- Homepage:
- Size: 517 KB
- Stars: 1
- Watchers: 5
- Forks: 0
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# knnp : Time Series Prediction using K-Nearest Neighbors Algorithm (Parallel)
First release was developed as an End-of-Degree Project.
Further improvements have been made now as a project from GRASIA investigation group:
https://grasia.fdi.ucm.es/
Purpose
----------
This package intends to provide R users or anyone interested in the field of time series prediction the possibility of aplying the k-nearest neighbors algorithm to time series prediction problems. Two main functionalities are provided:
- Time series prediction using this method.
- Optimization of parameteres *k* and *d* of the algorithm.
All the code involved has been optimized to:
- Parallelize critic components as the process of optimization of parameteres *k* and *d* or the calculation of distances.
- Use memory efficiently.
Authors
----------
- Daniel Bastarrica Lacalle
- Javier Berdecio Trigueros
Directors
----------
- Javier Arroyo Gallardo
- Albert Meco Alias
Maintainer
----------
- Daniel Bastarrica Lacalle
License
----------
AGPL-3