| Title |
Stochastinis dažnų posekių paieškos algoritmas / |
| Another Title |
Stochastic algorithm for mining frequent subsequences. |
| Authors |
Savulionienė, Loreta ; Sakalauskas, Leonidas |
| Full Text |
|
| Is Part of |
Jaunųjų mokslininkų darbai / Šiaulių universitetas.. Šiauliai : Šiaulių universiteto leidykla. 2011, nr. 4(36), p. 138-145.. ISSN 1648-8776 |
| Keywords [eng] |
Algorithms ; Database management ; Information science |
| Abstract [eng] |
The article introduces a stochastic algorithm for mining frequent subsequences, which randomly separates subsequences of different length when a database is being scanned. The distribution of the length of subsequences depends on a geometric law with parameter p and the distribution of the distance between chosen subsequences also depends on a geometric law with parameter q. The designed algorithm was tested using computer modelling including methods of statistical hypothesis testing, probability confidence limits, likelihood functions and Monte Carl0 for finding frequent subsequences. This algorithm is approximate, but it enables combining two important criteria (time and accuracy) and choosing values ofparameters p and q accordingly. Using analysis of random subsequences, the algorithm gives statistical conclusions about frequent subsequenees. Therefore, the designed stochastic algorithm for mining frequent subsequences can be used for search for frequent subsequences in large databases. |
| Published |
Šiauliai : Šiaulių universiteto leidykla |
| Type |
Journal article |
| Language |
Lithuanian |
| Publication date |
2011 |
