These are public posts tagged with #timeseries. You can interact with them if you have an account anywhere in the fediverse.
I had the pleasure of presenting this week at the @rladiesrome meetup about forecasting with linear regression. This workshop covered the following topics:Time series decomposition
Correlation and seasonal analysis
Modeling trend and seasonality
Using piecewise regression to model change in trend
Residuals analysis
The workshop recording is available online:
https://www.youtube.com/watch?v=lk3a3GQ7kc8
Enjoy the videos and music you love, upload original…
www.youtube.com Missed @ramikrispin @rladiesrome on using linear regression for forecasting time series? No worries! Catch the beginner-friendly session on YouTube
Enhance your skills now read our blog post and watch the recording: https://rladiesrome.org/talks/2025/meetup/07032025_RamiKrispin.html
#Rstats #Forecasting #timeseries
Beginner-friendly workshop led by author and open source…
R-Ladies Rome WebsiteThat means actively screening candidates for forecasting-specific red flags.
Don’t let red flag correlation make your company a casualty. In forecasting, red flags aren’t just signals—they’re causes.
These finding are as damming to time series LLMs as the “Are Transformers Effective for Time Series Forecasting?” was for transformers. It is hard to see how time series LLM will be able to deal with such hard evidence showing they don’t work.
A key research paper has now confirmed what many in the #TimeSeries community suspected: LLMs fundamentally fall short for forecasting tasks.
Is this surprising?
For those with significant experience in time series, the answer is a clear no.
FEDOT, да не тот
Привет, Хабр! Меня зовут Марина, я Head of Analytics and ML в SENSE , занимаюсь анализом данных уже более 5 лет. Сначала препарировала спектры в физике высоких энергий и сотрудничала с ЦЕРН-ом, а теперь строю рекомендательные системы и аналитику. В статье расскажу про опыт работы с пакетом FEDOT для прогнозирования временных рядов. Статья пригодится тем, кто хочет вкатиться в тему временных рядов и потыкать свои первые модельки на примере отечественных библиотек. Объясняю на примере задачи прогнозирования выходов кандидатов. Дисклеймер: во временных рядах я только начинаю свой путь, так что делюсь всеми своими фейлами и буду рада обратной связи в комментах.
https://habr.com/ru/companies/it_sense/articles/879048/
#Модели_временных_рядов #machinelearning #statistics #timeseries
Привет, Хабр! Меня зовут Марина, я Head of Analytics…
Хабр#programming #graphing #plotting #visualization #timeSeries #gnuplot #commonLisp #lisp #example https://screwlisp.small-web.org/programming/common-lisp-invoking-gnuplot/
I could not even find my own previous articles and #demos of this online!
I used #uiop run-program to handle one specific case like
(gnuplot "bad title" '((1 2) (3 4)) '((5 6) (7 8)))
or equivalently,
(apply 'gnuplot "bad title" '(((1 2) (3 4)) ((5 6) (7 8))))
Do you personally have an example? I remember it being hard to dredge up gnuplot examples but this is beyond silly.
I've been working with Time Series models a fair bit lately, and have noticed that ARIMA models (in particular) tend to regress to the long-term mean. This is usually a good thing, agreeing with the intuition that the least wrong guess about the future behaviour (of a stable system) is somewhere in that location (ie weather averages out to historical climate in the long run - in the absence of climate change, that is).
But it's made me wonder if these models might be the cause of consistent over|under-estimation of the future behaviour of systems undergoing fundamental changes over time?
eg Consistent underestimation of solar PV growth and overestimation of RBA wage-price figures.
New to time series forecasting?
Learn ARIMA modeling in this clear, hands-on guide from DataCamp!
Perfect for data science beginners. https://www.datacamp.com/tutorial/arima
#ARIMA #TimeSeries #DataScience #Forecasting
I am looking for a postdoc 'Environmental and behavioral health in a changing climate' (1/2)
2 years in #Rennes #france #rstats #timeseries
We are looking for a postdoctoral researcher to help us understand the short-term impacts of environmental conditions on mental health, sleep and physical activity related behaviors. Future findings will help us better anticipate present and future consequences of climate change on bike use and sleep.
Time Series Analysis with StatsModels
This workshop from the PyData Global 2024 conference by Allen Downey provides an introduction to time series analysis with the StatsModels Python library.
https://www.youtube.com/watch?v=foMbacbuAQk
#timeseries @python #DataScience #forecasting
Enjoy the videos and music you love, upload original…
www.youtube.com#Day6 of #30DayChartChallenge, theme: #FlorenceNightingale
Estacionalidad promedio de compraventas de vivienda en España (2007-2024). Coxcomb plot muestra meses con ventas típicamente >100 (altas) o <100 (bajas) vs media anual.
Método: X-13ARIMA-SEATS con #rstats {seasonal}. Datos: INE.
Código: https://t.ly/lC34V
Happy to share some updates to my package download forecasting analysis located here:
https://www.spsanderson.com/healthyverse_tsa/
I am now also using the amazing NNS package.
Thanks to @bmacs001, #OpenHistoricalMap has comprehensive coverage of #NewJersey municipality boundaries over time. Watch the state get divvied up into counties, townships, boroughs, cities, towns, and villages in this mesmerizing #TimeSeries:
Antti Rask and I worked on a package called #RandomWalker which seems to have download patterns that are anything but. There is a lot of work lined up to do on it, time is elusive but we will get there.
You can follow the issues here: https://github.com/spsanderson/RandomWalker/issues
Instead, we get over-engineered LLMs struggling with fundamentals that classical methods mastered decades ago.
Maybe it’s time to put R&D dollars where they actually move the needle. #TimeSeries #AI #TechRants
#Forecasting#DataScience #ProfessionalDevelopment hashtag#Analytics #Learning #timeseries #machinelearning #nohype
Simply search for "Forecasting Mastery" on maven.com and take the first step toward transforming your forecasting skills.
Don’t miss this chance to elevate your expertise—enroll today and never look back!
https://maven.com/valeriy-manokhin/modern-forecasting-mastery
Gain hands-on experience with cutting-edge forecasting…
maven.com