Archives ISF - Open Forecasting

ISF2024: How to Bootstrap Time Series without Attracting Attention of Statisticians

Ivan Svetunkov — Wed, 03 Jul 2024 14:11:41 +0000

On 1st July, I presented my ongoing work on time series bootstrap and its impact on prediction intervals at ISF2024 in Dijon, France.

Abstract: Bootstrap is extensively used in statistics and machine learning for cross-sectional data to account for uncertainty about the data, model form, and parameter estimates. However, conventional methods may not be suitable for time series data due to autocorrelation and specific dynamic structures. Over the years, various approaches have been developed to address this issue. Some assume specific models (e.g., STL), while others are non-parametric (e.g., Maximum Entropy Bootstrap, MEB). However, the former can be overly restrictive, while the latter may not perform well in case of outliers and external drivers. To address these issues, we propose a non-parametric bootstrap approach inspired by MEB, which does not assume any structure in the data yet creates reasonable copies of existing time series of different nature. These copies can be utilised in bagged ETS/ARIMA or any other approach involving small sample uncertainty. We demonstrate how the proposed bootstrap works using real-time series examples and assess improvements it brings in terms of forecasting accuracy compared to conventional approaches.

Here are the slides of the presentation.

And here is me, trying not to attract attention of statisticians:

How to attract attention of statisticians…

Message ISF2024: How to Bootstrap Time Series without Attracting Attention of Statisticians first appeared on Open Forecasting.

ISF2022: How to make ETS work with ARIMA

Ivan Svetunkov — Wed, 20 Jul 2022 12:06:48 +0000

This time ISF took place in Oxford. I acted as a programme chair of the event and was quite busy with schedule and some other minor organisational things, but I still found time to present something new. Specifically, I talked about one specific part of ADAM, the part implementing ETS+ARIMA. The idea is that the two models are considered as competing, belonging to different families. But we have known how to unite them at least since 1985. So, it is about time to make this brave step and implement ETS with ARIMA elements.

ETS+ARIMA love story with happy ending…

This talk was based on Chapter 9 of ADAM monograph, and more specifically on Section 9.4.

The slides of the presentation are available here.

Message ISF2022: How to make ETS work with ARIMA first appeared on Open Forecasting.

ISF2021: How to Make Multiplicative ETS Work for You

Ivan Svetunkov — Wed, 30 Jun 2021 12:09:46 +0000

This year International Symposium on Forecasting was held online, although Centre for Marketing Analytics and Forecasting of Lancaster University had their own hub, where we would come and watch presentations together and even present to the others.

Presentation at ISF2021 Lancaster Hub

I presented on the topic of Multiplicative ETS, based on this chapter of the ADAM textbook and on the paper John Boylan and I are working on. In this presentation, I discuss that the point forecasts from the ETS(M,*,*) models in general do not correspond to conditional expectations (and not even to geometric means) and show the difference between the two. Furthermore, we propose using Log Normal, Gamma and Inverse Gaussian distributions for the error term of the model instead of the Normal one. This makes the model more realistic and help in forecasting on low volume data. All of this is already implemented in adam() function of smooth package for R.

Here are the slides.

Message ISF2021: How to Make Multiplicative ETS Work for You first appeared on Open Forecasting.

International Symposium on Forecasting 2019

Ivan Svetunkov — Wed, 03 Jul 2019 09:12:25 +0000

The ISF2019 took place in Thessaloniki, Greece. This time I presented a spin-off of my research on intermittent demand in retail, entitled as “What about those sweet melons? Using mixture models for demand forecasting in retail”. The idea is quite trivial and simple: use mixture distribution regressions (e.g. logistic and log-normal distributions) in order to predict the seasonally-intermittent sales in retail. The model is quite simple and easy to implement in practice. The main problem that I’ve faced so far is the absence of the proper data. I only had 24 series of weekly sales of tomatoes provided by a small company, but I need more in order to see, which of the approaches works best. For this research, I need the data like this:

Retail sales of tomatoes

Until I have the data, I cannot write a paper on that topic…

Anyway, here are the slides if anyone wants to have a look.

Message International Symposium on Forecasting 2019 first appeared on Open Forecasting.

International Symposium on Forecasting 2018

Ivan Svetunkov — Tue, 19 Jun 2018 23:12:42 +0000

This year I have presented an extension of the research from ISF2017, called “Forecasting intermittent data with complex patterns”. This time we developed the model with “logistic probability”, which allows capturing complex patterns in demand occurrence part of the data. I also tried making the presentation more entertaining and easier to understand by a wider audience. But don’t worry, I did not abandon formulae, this was still a proper math-based presentation.

Here are the slides from the presentation. My colleagues and I are working on a paper for this. But this is related to the paper John and I submitted to IJF.

An example of time series discussed in the presentation

There is also now a video of my presentation:

Message International Symposium on Forecasting 2018 first appeared on Open Forecasting.

International Symposium on Forecasting 2017

Ivan Svetunkov — Sun, 30 Jul 2017 11:27:43 +0000

In Cairns, Australia, I have presented the topic from Bath — One for all: forecasting intermittent and non-intermittent demand using one model. It was well received although one of the professors did not understand the main point of the presentation and I could not explain him, why this is important and why the proposed approach is useful. Oh, well. That’s life!

Message International Symposium on Forecasting 2017 first appeared on Open Forecasting.

International Symposium on Forecasting 2016

Ivan Svetunkov — Fri, 24 Jun 2016 22:28:25 +0000

This time I presented a talk on Trace Forecast Likelihood, based on the presentation given at Ghent and in Higher School of Economics earlier this year. Unfortunately, I was in the session of statisticians, who discussed hypothesis testing and weren’t aware of the area of my presentation. As a result, this presentation passed without any comments and feedback. Still, this was a great conference with a lot of interesting talks and interesting people.

Message International Symposium on Forecasting 2016 first appeared on Open Forecasting.

International Symposium on Forecasting 2015

Ivan Svetunkov — Thu, 25 Jun 2015 14:28:29 +0000

I have given a presentation on ISF2015 in Riverside, USA. The topic of presentation was Complex Exponential Smoothing for Time Series Forecasting. This is based on a paper submitted to International Journal on Forecasting, which is currently under review. Here’s the presentation.

Message International Symposium on Forecasting 2015 first appeared on Open Forecasting.