The nnfor (development version here) package for R facilitates time series forecasting with Multilayer Perceptrons (MLP) and Extreme Learning Machines (ELM). Currently (version 0.9.6) it does not support deep learning, though the plan is to extend this to this direction in the near future. Currently, it relies on the neuralnet package for R, which provides… Read More: Tutorial for the nnfor R package »
A feature that was missing in the nnfor package was the ability to reuse and retrain mlp and elm model fits. This is now possible with the new arguments model and retrain. As an example, let us use the AirPassengers time series, with three different sample sizes and re-use and re-train the same model in… Read More: Update for nnfor: reuse and retrain models »
I have been asked a few time to provide an example how to use MAPAx, from the MAPA package for R, so I prepared this blog post. I admit the documentation could be better, so I put together this example from a retailing case – the original setting MAPAx was developed for (see paper here).… Read More: MAPAx example for R »
I have put up a github repository for the nnfor package for R: https://github.com/trnnick/nnfor I will be putting updates and fixes there, before they are pushed on CRAN. You can also report there bugs. You can install the current github version with: Related PostsPrinciples of Business Forecasting 2e Benchmarking Facebook’s Prophet Forecasting time series with… Read More: nnfor on github »
You can find the material from yesterday’s workshop at the International Symposium on Forecasting, 2017 here. The workshop notes assume some knowledge of what the various forecasting methods do, and the main focus is on showing which functions to use and how, so as to perform a wide variety of forecasting tasks: Time series exploration… Read More: Material for `Forecasting with R: A practical workshop’ at ISF2017 »
The previous version of the MAPA package implemented only the univariate aspect of the algorithm. Version 2.0.1 implements MAPAx as well, which allows incorporating regressors in your forecasts. In this paper we demonstrated the usefulness of temporal aggregation in the case of forecasting demand in the presence of promotions. In particular, we showed that MAPAx… Read More: MAPAx available for R & new MAPA package on CRAN »
Here is the link: https://github.com/trnnick/mapa It has been a long time I wanted to rework the MAPA package for R, but I could not find the time. Finally I got around starting it. There are three objectives in this: Clean up code and introduce S3methods. MAPA was the first package that I wrote! Incorporate the… Read More: MAPA package for R on GitHub »
We have been re-working the TStools package over the past couple of weeks. The major changes are: The es function that is an alternative to the ets function from the forecast package has been removed. Now it is published separately in the smooth package, which contains a collection of interesting implementations for exponential smoothing, ARIMA… Read More: TStools recent changes »
Together with Fotios Petropoulos we gave a workshop on producing forecasts with R, at the International Symposium on Forecasting, 2016. You can find the material of the workshop here. The workshop notes assume knowledge of what the various forecasting methods do, which are only briefly explained in the workshop’s slides, and mostly focuses in showing… Read More: Material for `Forecasting with R: A practical workshop” »
Version 1.8 of tsintermittent has been submitted to CRAN and should be shortly available for download. Amongst various new checks on inputs to better accommodate handling multiple time series with data frames, a new option has been added to data.frc. When method=”auto” two things will happen: Function idclass(…,type=”PKa”) will be called to classify the time… Read More: Another update for tsintermittent »