> ML_ALGORITHM // HIDDEN-MARKOV-MODELS-BAUM-WELCH_v1.0
Hidden Markov Models (HMM / Baum-Welch / Viterbi)
Generative sequence model modeling observed temporal sequences as emissions from an underlying unobservable finite-state Markov chain.
Probabilistic Graphical Modelsbayesian-probabilistichigh-intrinsicsmall (<1k)
Back to All AlgorithmsComputational Complexity
Training Complexity:O(iterations * T * |S|^2) via Baum-Welch
Inference Complexity:O(T * |S|^2) via Viterbi decoding
Hardware Profile
CPU Friendly:Yes
Requires GPU:No
Memory Footprint:low
Interpretability & Data
Interpretability Tier:high-intrinsic
Training Data Needs:small (<1k)
Interpretability Assessment
Viterbi algorithm outputs the mathematically optimal global sequence of hidden regime states.
Suitable Tasks & Supported Modalities
Suitable Tasks:
sequence labelingspeech phoneme recognitionfinancial regime detection
Supported Modalities:
time-seriestextaudio
Implementing Libraries
hmmlearn
SciPySciPy Community / NumFOCUS · v1.14.1
View SpecFoundational Literature
Common Pitfalls & Warnings
- First-order Markov assumption cannot capture long-range temporal dependencies spanning multiple timesteps