Professional Documents
Culture Documents
Where: P (S0) Specifies Initial Conditions P (St+1 - ST) Specifies The Dynamics P (Ot - ST) Specifies The Sensor Model
Where: P (S0) Specifies Initial Conditions P (St+1 - ST) Specifies The Dynamics P (Ot - ST) Specifies The Sensor Model
Hidden Markov Models are and how they are used to reason under uncertainty.
Your response should discussion the role of probability distributions and smoothing.
According to Wikipedia, Hidden Markov model (HMM) is a statistical Markov model
in which the system being modeled is assumed to be a Markov process with
unobserved (hidden) states. An HMM can be presented as the simplest dynamic
Bayesian network. (Wikipedia, 2014). Hidden Markov Models are a ubiquitous tool
for modelling time series data or to model sequence behavior. They are used in
almost all current speech recognition systems. The state of the process in HMMs
model is described by a Single Discrete Random Variable. Possible States of the
World is the possible values of the variable in HMMs. While staying within the HMM
network, the additional state variables can be added to a temporal model. To solve
temporal probabilistic reasoning, HMM is used, independent of transition and
sensor model.
A Hidden Markov Model (HMM) is a belief network according to (Poole &
Mackworth, 2010): Consider the figure below:
Where
P(S0) specifies initial conditions
P(St+1|St) specifies the dynamics
P(Ot|St) specifies the sensor model
Now,
Suppose a robot wants to determine its location based on its actions and its
sensor readings. Called Localization
This can be represented by the augmented HMM as shown below:
Applications