The R Package knnwtsim: Nonparametric Forecasting With a Tailored\n Similarity Measure

2021/12/12 by Matthew Trupiano, Trupiano, Matthew
Computer Science · Decision Sciences · #Computation (stat.CO) #Data Analysis with R #FOS: Computer and information sciences #Forecasting Techniques and Applications #Methodology (stat.ME)

paper · pdf · doi:10.48550/arxiv.2112.06266

openalex publication_date 2021/12/12 · openalex created_date 2022/05/05 · openalex updated_date 2026/07/28

Abstract

The R package knnwtsim provides functions to implement k nearest neighbors\n(KNN) forecasting using a similarity metric tailored to the forecasting problem\nof predicting future observations of a response series where recent\nobservations, seasonal or periodic patterns, and the values of one or more\nexogenous predictors all have predictive value in forecasting new response\npoints. This paper will introduce the similarity measure of interest, and the\nfunctions in knnwtsim used to calculate, tune, and ultimately utilize it in KNN\nforecasting. This package may be of particular value in forecasting problems\nwhere the functional relationships between response and predictors are\nnon-constant or piece-wise and thus can violate the assumptions of popular\nalternatives. In addition both real world and simulated time series datasets\nused in the development and testing of this approach have been made available\nwith the package.\n

Related

• Applying k-nearest neighbors to time series forecasting : two new approaches
• Nearest Neighbor Multivariate Time Series Forecasting
• G-NM: A Group of Numerical Time Series Prediction Models
• Time Series Analysis and Modeling to Forecast: a Survey
• Forecasting with sktime: Designing sktime's New Forecasting API and Applying It to Replicate and Extend the M4 Study