A More Fluid Way To Model Time-Series Data
IBM, September 30,2025
IBM's new time-series foundation model, FlowState, uses a state-space architecture to outperform much larger models on an industry leaderboard known for its challenging mix of short and long-term forecasting problems.
The past is often a powerful guide to the future, whether you're trying to predict stock prices, traffic on a network, or quarterly sales. But so much depends on the data.
Time-series data can be sliced in slivers of seconds and minutes, or in chunks spaced out over days and years - and sometimes even at different rates within the same sample. Things quickly get complicated once you move from modeling a time series captured at one scale to modeling multiple timescales at once, as foundation models are designed to do.