Atlas record: https://www.aitimeline.dev/en/events/evt-0183/
Research · February 1970
The Baum-Welch algorithm
Baum and co-authors described how to estimate the parameters of a hidden Markov process from observations, without ever seeing the states.
Why it matters
A model where the cause is hidden and only the effect is visible became trainable, and that defined speech recognition for thirty years.
The paper is purely mathematical and never mentions speech. But it gives the iterative procedure that raises likelihood at every step, later recognised as a special case of expectation maximisation. A speech signal fits the model well: articulation is hidden and only acoustics are observed. Without this result neither DRAGON nor the IBM systems would have had anything to train.
Event record
- Event date
- February 1970
- Timeline date
- Event date
- Verification
- Sources gathered automatically · September 17, 2026
- Lines
- ID
- evt-0183
The February 1970 issue of the Annals of Mathematical Statistics.
Related events
Earlier
- Related Schlesinger: unsupervised learning as repeated learning 1968links to this record
Both estimate parameters without seeing the hidden labels or states, by iterations that raise the likelihood: Schlesinger for a mixture of distributions in recognition, Baum and co-authors for a hidden Markov process. The record claims no causal link between them.
- Related Dynamic Programming 1957
The estimation rests on the same step-by-step decomposition as dynamic programming.
Later