Home > Search Results: Katuščák, S
Search:
Search Tips :: Advanced Search
Search collections:
Sort by: Display results: Output format:
National Repository of Grey Literature 2 records found  Search took 0.00 seconds. 
Software for Bidirectional Sequence Classification
 Katuščák, Slavomír ; Raab, Jan (advisor) ; Spoustová, Johanka (referee)
The goal of this master thesis was to implement algorithm for bidirectional classifi cation of sequences described in article of Libin Shen and col. This work builds on implementation and experimental project Mor ce, morphological tagger based on hidden Markov model with averaged perceptron and on Jan Raab's master thesis, which goal was to implement Viterbi's algorithm and nd best possible set of features for it. This work o ers Shen's algorithm as alternative to Viterbi's algorithm. Limitation of Viterbi's algorithm is in its one-way direction, selection of tag on some position can depend on some already done decisions, but only on preceeding positions. This can cause selection of suboptimal candidate and regression of analyser accuracy. This work implements algorithm, that works bidirectionally and can use already done decisions from both sides of current position, keeping tolerable complexity.
Detailed record
Software for Bidirectional Sequence Classification
 Katuščák, Slavomír ; Raab, Jan (advisor) ; Spoustová, Johanka (referee)
The goal of this master thesis was to implement algorithm for bidirectional classifi cation of sequences described in article of Libin Shen and col. This work builds on implementation and experimental project Mor ce, morphological tagger based on hidden Markov model with averaged perceptron and on Jan Raab's master thesis, which goal was to implement Viterbi's algorithm and nd best possible set of features for it. This work o ers Shen's algorithm as alternative to Viterbi's algorithm. Limitation of Viterbi's algorithm is in its one-way direction, selection of tag on some position can depend on some already done decisions, but only on preceeding positions. This can cause selection of suboptimal candidate and regression of analyser accuracy. This work implements algorithm, that works bidirectionally and can use already done decisions from both sides of current position, keeping tolerable complexity.
Detailed record

See also: similar author names
2 Katuščák, Slavomír
Interested in being notified about new results for this query?
Subscribe to the RSS feed.
Digital Repository :: Search :: Submit :: Personalize :: Help
Powered by Invenio v1.1.2
Maintained by admin.invenio@techlib.cz
This site is also available in the following languages:
Česky  English