Handle forecast uncertainty with StatsForecast prediction intervals!
Point forecasts provide only a single predicted value without any uncertainty information. You can't assess how confident the model is in its prediction, making risk assessment impossible.
StatsForecast…
(Shapley) Explanations are now available in NeuralForecast v3.1.0! 🎉
The latest release adds SHAP-powered explanations to univariate forecasting models in NeuralForecast, making it clear which features drive each forecast via the explain() method.
The feature contribution…
A new benchmark tests 9 models across 21 crypto assets on daily and hourly data. Two standout results:
🔹 Accuracy: Fine-tuned TimeGPT (no variables) leads on average across daily and hourly
🔹 Speed: Zero-shot TimeGPT is 10 times faster than deep nets, enabling rapid iteration…
When AI in healthcare is discussed, the focus often goes to diagnostics or patient chatbots. While those are valuable, many of the most consistent cost pressures arise in operations and supply chains.
For example:
🔹 Stockouts can lead to emergency orders at higher prices.
🔹…
We’re proud to be ranked by G2, the world’s largest software marketplace, where real users review and rank business tools, as:
🏆 Highest Performer
🍰 Easiest to Use
Thanks to everyone who has supported us on this journey to make time series forecasting faster, more accurate,…
Eliminate target transformation errors with MLforecast (automated preprocessing)!
Time series forecasting often requires target variable transformations. For example, you may want to apply a difference transformation to make the target variable stationary.
But relying on manual…
Want to land a job as a Machine Learning Engineer at OpenAI or DoorDash?
Turns out brushing up on your Python isn’t enough…
Step 1: Know TimeGPT 😉 (it’s literally in their job descriptions)
Step 2: Apply
Step 3: Show off those forecasts 🚀
Discover which automated time series features drive model performance with MLforecast!
Which features from your automated time series engineering actually improve predictions? Without visibility into feature importance, you don't know whether your automated feature engineering…
Generate statistical lag features with one dictionary in MLforecast!
Lag features are time-shifted versions of your data that capture temporal dependencies, enabling models to learn how past values influence future predictions.
Building these features manually requires complex…
Access 7 specialized long-horizon forecasting datasets for extended predictions!
Long-horizon datasets enable researchers to validate model performance across multiple years and changing market conditions.
These extended datasets support comprehensive benchmarking that reveals…
Scale time series forecasting across any DataFrame format!
Most time series libraries lock you into a single DataFrame framework, forcing painful migrations when scaling up.
Switching frameworks usually means rewriting data loading, preprocessing, and model interfaces with tons…
Transform forecasting workflows with TimeGPT's Polars support!
Large time series datasets cause memory overflow when you try to load everything into pandas for forecasting.
Your workflow crashes before you even get to model training, wasting hours of preprocessing work.…
Enhance anomaly detection with TimeGPT exogenous features!
Do you find your anomaly detection models flagging expected seasonal events as outliers?
Without context about holidays, events, or calendar patterns, models treat predictable spikes as anomalies.
TimeGPT supports…
What if you could improve your forecasts by 35%… in 2 weeks?
That’s exactly what happened when @Decathlon teamed up with Nixtla.
Using TimeGPT, the world’s largest sporting goods retailer:
📈 Improved accuracy by 35%
📉 Reduced bias by 77%
⚡ Unified millions of time series…
36K Followers 5K FollowingExperienced Data Science Leader | PhD in Machine Learning | 4x Author | Black Belt 🥋 in Time Series | Chief Conformal Prediction Promoter| Mathematician |
28K Followers 599 FollowingLLMs and retrieval by day and other genres of AI when I get the chance
🧪 Senior AI Eng @NVIDIAAI
🏫 @fastdotai trained DL Eng
📝 https://t.co/By87iXx5Pu
62 Followers 672 FollowingGeomatics and Geophysics Engineer. Applied AI researcher. MSc in Seismology and is an avid enthusiast of Astrophysics. PhD Candidate. Orgullo FI UNAM
397 Followers 4K FollowingPhilosophy, math, and CS.
Presence beats reactivity. Joy beats ego. Truth sets you free.
Curious about consciousness, the cosmos, and silence.
36K Followers 5K FollowingExperienced Data Science Leader | PhD in Machine Learning | 4x Author | Black Belt 🥋 in Time Series | Chief Conformal Prediction Promoter| Mathematician |
1.3M Followers 1K FollowingCo-Founder of Coursera; Stanford CS adjunct faculty. Former head of Baidu AI Group/Google Brain. #ai #machinelearning, #deeplearning #MOOCs
69K Followers 2K Following✨ AI should be about empowering humans, building understanding, and making dreams realities. 👩💻 DevX Eng. Lead @GoogleDeepMind ex-@GitHub || views = my own!
79K Followers 4K FollowingConnect, Unify, Respond to any data, anywhere with human-level intelligence.
@Forbes Most Promising AI Companies, @Gartner_inc Cool Vendor
316K Followers 3K FollowingWWCode Summit at TechFutures Oct. 7-9
Get 🎟️ at https://t.co/ToXlNzohMe
⚡️WWCode is BACK!
⚡️500K+ technologists, 147+ Countries, leveling up our skills #wwcode
2K Followers 301 Following#Forecasting insight from top academics and business practitioners in the field. Published by the nonprofit International Institute of Forecasters.
4K Followers 2K FollowingContinuously trying to figuring it out. Doing: Startup stuff for @PyCon et al. ⚡ Prior adventures: 🧠 @DeepgramAI 🔎📈 @CrunchbaseNews, et al.
255 Followers 950 Followingprofessional husbande / ai / econ / longevity.
no one has enemies, only nakamas. embrace sintropy
techno-distributed-optimization
730 Followers 833 FollowingData Scientist | Time Series Forecasting | Machine Learning | @UniMelb Research Fellow in Data Science @MelbCtrDataSci | Forecasting Analytics @EnergyAustralia
421 Followers 100 FollowingI was bought in a shop... with a discount.
Work in Lancaster Centre for Marketing Analytics and Forecasting, Lancaster University
85 Followers 57 FollowingProgramme d'accélération pour startups en IA basé à Montréal.
English version : @TechstarsMTL
Alumni : https://t.co/AzT2MluV4Y
9K Followers 462 FollowingNow @tigerdatabase. The modern cloud platform built on PostgreSQL for time series, events, and analytics (and vectors too). ⭐️ - https://t.co/9HK3eQGIr5.
206 Followers 119 FollowingAccelerator focused on advancing the commercial applications of AI.
En français: @techstars_mtlFR
Alumni: https://t.co/JbNPXVs6cK