MOMENT, a recently released family of foundation models, aims to tackle time series forecasting, classification, and more; Nikos Kafritsas explains how it works, unpacks its architecture, and compares its performance to other state-of-the-art models. buff.ly/3xRrKka

In his latest exploration of new models, Nikos Kafritsas explains the inner workings of MOMENT, a recently released foundation model for time series forecasting, classification, and anomaly detection. buff.ly/3xRrKka

𝐌𝐎𝐌𝐄𝐍𝐓: 𝐀 𝐅𝐨𝐮𝐧𝐝𝐚𝐭𝐢𝐨𝐧 𝐌𝐨𝐝𝐞𝐥 𝐟𝐨𝐫 𝐓𝐢𝐦𝐞 𝐒𝐞𝐫𝐢𝐞𝐬 𝐅𝐨𝐫𝐞𝐜𝐚𝐬𝐭𝐢𝐧𝐠, 𝐂𝐥𝐚𝐬𝐬𝐢𝐟𝐢𝐜𝐚𝐭𝐢𝐨𝐧, 𝐀𝐧𝐨𝐦𝐚𝐥𝐲 𝐃𝐞𝐭𝐞𝐜𝐭𝐢𝐨𝐧 𝐚𝐧𝐝 𝐈𝐦𝐩𝐮𝐭𝐚𝐭𝐢𝐨𝐧

aihorizonforecast.substack.com

Thank you for sharing!

MOIRAI: Salesforce’s Foundation Model for Time-Series Forecasting - Code, model weights, and data will be released soon by Nikos Kafritsas buff.ly/3ThTqFF

To learn how TimesFM — Google's new model for time-series forecasting — works, don't miss Nikos Kafritsas's latest explainer, which makes the paper's essential components accessible for a broader audience of ML practitioners. buff.ly/3SVWpDM

Stay up-to-date with the latest foundation models arriving on the scene: Nikos Kafritsas's new paper walkthrough unpacks the inner workings of TimesFM, Google's new model for time-series forecasting. buff.ly/3SVWpDM

𝐀𝐈 𝐇𝐨𝐫𝐢𝐳𝐨𝐧 𝐅𝐨𝐫𝐞𝐜𝐚𝐬𝐭 #3 𝐢𝐬 𝐨𝐮𝐭!

➡️ The 3rd edition describes 𝗧𝗶𝗺𝗲𝘀𝗙𝗠, Google's new 200B Foundation Model for TS Forecasting.

➡️ Contains an updated tutorial on 𝐓𝐞𝐦𝐩𝐨𝐫𝐚𝐥 𝐅𝐮𝐬𝐢𝐨𝐧 𝐓𝐫𝐚𝐧𝐬𝐟𝐨𝐫𝐦𝐞𝐫.

Link: aihorizonforecast.substack.com/p/timesfm-goog…

𝐑𝐮𝐩𝐭𝐮𝐫𝐞𝐬: 𝐂𝐡𝐚𝐧𝐠𝐞 𝐩𝐨𝐢𝐧𝐭 𝐝𝐞𝐭𝐞𝐜𝐭𝐢𝐨𝐧 𝐥𝐢𝐛𝐫𝐚𝐫𝐲 (𝐏𝐲𝐭𝐡𝐨𝐧)

For latest developments in time-series research, subscribe to my newsletter:

AI Horizon Forecast newsletter: aihorizonforecast.substack.com
Ruptures docs: centre-borelli.github.io/ruptures-docs/

.Nikos Kafritsas explores the AutoGluon-Timeseries (AG-TS), outlines its capabilities, and constructs a simple project utilizing a widely known Tourism dataset: buff.ly/3S4N11k

AutoGluon-TimeSeries: Every Time Series Forecasting Model In One Library by Nikos Kafritsas in Towards Data Science towardsdatascience.com/autogluon-time…

AutoGluon-TimeSeries: Every Time Series Forecasting Model In One Library by Nikos Kafritsas in Towards Data Science
Knut Jägersberg Chidambara .ML. Kevin Corella N. 🔬🔭 🇪🇺
Eveline Ruehlin Estela Mandela Dr. de l'immatériel achutha subhash Deus ex Machina BusinessIntelligence

towardsdatascience.com/autogluon-time…

AutoGluon-TimeSeries: Every Time Series Forecasting Model In One Library - A powerful library by Amazon  (coding example included) by Nikos Kafritsas buff.ly/3S4N11k

Wow!

This is a snippet from the 69-page filing of Time's lawsuit against OpenAI and Microsoft.

ChatGPT has memorized entire articles from The New York Times, some of which are paywalled.

Left column: ChatGPT output
Right column: Actual NYT article

Link : nytco-assets.nytimes.com/2023/12/NYT_Co…

𝐀𝐮𝐭𝐨𝐆𝐥𝐮𝐨𝐧 - 𝐓𝐢𝐦𝐞𝐒𝐞𝐫𝐢𝐞𝐬 (𝐀𝐆-𝐓𝐒) : A powerful time-series forecasting framework by Amazon:

➡️Offers all SOTA forecasting models (statistical, ML, Deep Learning)

➡️ Leverages ensembling

➡️ Open-Source

Link:aihorizonforecast.substack.com/p/autogluon-ti…

𝐀𝐩𝐩𝐥𝐞 𝐡𝐚𝐬 𝐣𝐮𝐬𝐭 𝐫𝐞𝐥𝐞𝐚𝐬𝐞𝐝 𝐌𝐋𝐗, 𝐚𝐧 𝐨𝐩𝐞𝐧-𝐬𝐨𝐮𝐫𝐜𝐞 𝐃𝐞𝐞𝐩 𝐋𝐞𝐚𝐫𝐧𝐢𝐧𝐠 𝐥𝐢𝐛𝐫𝐚𝐫𝐲 𝐟𝐨𝐫 𝐀𝐩𝐩𝐥𝐞 𝐒𝐢𝐥𝐢𝐜𝐨𝐧 𝐚𝐧𝐝 𝐌𝐚𝐜𝐎𝐒

The syntax of MLX is very similar to PyTorch:

Code: github.com/ml-explore/mlx
Docs:ml-explore.github.io/mlx/build/html…

Time-Series Forecasting: Deep Learning vs Statistics — Who Wins? by Nikos Kafritsas in Towards Data Science towardsdatascience.com/time-series-fo…