TY - GEN
T1 - Discovery of Rare Itemsets Using Hyper-Linked Data Structure
T2 - 13th International Conference on Applications and Techniques in Information Security, ATIS 2022
AU - Yadavalli, Goutham
AU - Rai, Shwetha
N1 - Publisher Copyright:
© 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
PY - 2023
Y1 - 2023
N2 - Pattern mining has been more important in the solution of various data mining jobs over the years. The extraction of common patterns was the primary focus of pattern mining research for a long period of time, with the mining of rare patterns being neglected. Rare pattern mining is becoming more popular as researchers recognize the importance of rare patterns. The hyper-linked data structure is suitable to store sparse data set in the main memory and enables dynamic adjustment of links during the mining process using recursion. However, a sequential approach to discovering rare patterns from a large dataset is inefficient. Hence a CUDA-based parallel algorithm has been implemented to discover rare itemsets. The algorithm is tested using dense and sparse datasets on a GPU. The GPU initialization time affects the time taken to discover rare itemsets. The time taken to transfer data between CPU and GPU is significantly large and the parallel implementation of an algorithm with a recursive approach is unsuitable.
AB - Pattern mining has been more important in the solution of various data mining jobs over the years. The extraction of common patterns was the primary focus of pattern mining research for a long period of time, with the mining of rare patterns being neglected. Rare pattern mining is becoming more popular as researchers recognize the importance of rare patterns. The hyper-linked data structure is suitable to store sparse data set in the main memory and enables dynamic adjustment of links during the mining process using recursion. However, a sequential approach to discovering rare patterns from a large dataset is inefficient. Hence a CUDA-based parallel algorithm has been implemented to discover rare itemsets. The algorithm is tested using dense and sparse datasets on a GPU. The GPU initialization time affects the time taken to discover rare itemsets. The time taken to transfer data between CPU and GPU is significantly large and the parallel implementation of an algorithm with a recursive approach is unsuitable.
UR - https://www.scopus.com/pages/publications/85161107071
UR - https://www.scopus.com/inward/citedby.url?scp=85161107071&partnerID=8YFLogxK
U2 - 10.1007/978-981-99-2264-2_23
DO - 10.1007/978-981-99-2264-2_23
M3 - Conference contribution
AN - SCOPUS:85161107071
SN - 9789819922635
T3 - Communications in Computer and Information Science
SP - 290
EP - 301
BT - Applications and Techniques in Information Security - 13th International Conference, ATIS 2022, Revised Selected Papers
A2 - Prabhu, Srikanth
A2 - Pokhrel, Shiva Raj
A2 - Li, Gang
PB - Springer Science and Business Media Deutschland GmbH
Y2 - 30 December 2022 through 31 December 2022
ER -