A study of minimum classification error training for segmental switching linear Gaussian hidden Markov models.
2004
A system and method for facilitating a telecommunication subscriber connection using a domain name system is disclosed. The system, which is part of a DNS server, supports connection to a mobile terminal through a data network and includes multiple connections to external devices or systems, although one or more of the connections may be serviced by a single link. The connections link the system to the data network, to a home location register ("HLR") associated with the terminal and to a mobile switch associated with the terminal. Upon initiation of a query to the DNS server through the PDN, the DNS server signals the HLR for the identification of the switch that is associated with the mobile terminal. The HLR returns an identifier for the switch and the DNS server then requests the switch to provide an address for the mobile terminal. The switch establishes a connection with the mobile terminal and provides a temporary address through a network interface system such as an interworking function. The switch then returns this new address to the DNS server.
Keywords:
- Network interface
- Artificial intelligence
- Pattern recognition
- Hidden semi-Markov model
- Markov model
- Computer science
- Dynamic Bayesian network
- Identifier
- Hidden Markov model
- Variable-order Markov model
- Domain Name System
- Speech recognition
- Gaussian
- Causal Markov condition
- Variable-order Bayesian network
- Maximum-entropy Markov model
- Correction
- Source
- Cite
- Save
- Machine Reading By IdeaReader
8
References
7
Citations
NaN
KQI