Conformal prediction for #timeseries#forecasting is the hottest π₯π₯π₯π₯ ππππ area of research and applications right now.
The whole story started in summer 2021 when researchers from Georgia Tech published the first ever paper applying conformal prediction to time series.
and presenting it at the first event at major conference totally dedicated to conformal prediction (ICML DFUQ 2021 organised by @ml_angelopoulos and @stats_stephen
Roll forward just 1.5 yeast - there are multiple papers published on the subject, several #opensoure libraries like MAPIE implemented the original first ever paper from Georgia Tech (method called EnbPI).
Many well known companies are implementing conformal prediction for time series forecasting to provide significant boost to business value delivered by data science teams
The new paper from the originals in time series forecasting research (Georgia Tech team) was just published.
βSequential Predictive Conformal Inference for Time Seriesβ significantly improves the original model.
@Zergylord@SchmidhuberAI Personally I find it hilarious that DeepMind is turning phd interns whilst some of its grunts can engage in trolling on Twitter during working time.
@Zergylord@SchmidhuberAI Cost saving on interns but it is ok apparently to troll for hours on Twitter during working day π
Motivated by having seen yet another Plattβs scaler post.
Plattβs scaling and isotonic regression are ~20 years old at this point. Both of them donβt have any mathematical guarantees of validity and are outperformed by conformal prediction Venn-ABERs
VENN-ABERS is in fact a better regularised version of isotonic regression that constructors two isotonic regressions by postulating that a test object can a priori have both 0 and 1 as a label.
By doing that VENN-ABERS is able to achieve theoretical guarantees of validity (lack of bias) as the expense of multi probability prediction where the test object will have two probabilities of class 1 instead of one. p0 is a lower bound and p1 is an upper bound for class 1 prob
Last year a few data scientists including myself exposed severe issues with facebook prophet resulting in facebook pulling down claims such as βwith facebook prophet anyone can produce better forecasts than human experts β.
Still the hope was that with Neural Prophet things will be better and one can at least partially repair the worst forecasting model of the XXIst century - Facebook Prophet. valeman.medium.com/benchmarking-fβ¦