Parameter setting procedure via quick parameter evaluation in frequent pattern mining for outbreak detection

Abstract

Data sources for outbreak detection nowadays not only focus on emergency department or hospital-based data but also grocery data. However, the size of huge data, may consume higher time and extreme number of discovered pattern. Unfortunately not all the discovered pattern from the frequent mining is interesting pattern. Hence frequent pattern mining algorithms producing numbers of frequent pattern, still parameter uses in minimum support and which frequent itemset producing better pattern remains fairly open. It is important to gains some limitation of minimum support to be applied to the frequent mining algorithm so that we not end up at compiling higher patterns including a normal pattern. We propose a procedure based on quick parameter setting to estimate minimum support and also frequent itemset. Our empirical validation shown the procedure will extract ranging minimum support and frequent itemset to be considered to generate interesting pattern.Keywords-frequent pattern mining, minimum support, frequent itemset.

Frequent Pattern Mining and Current State of the Art

“…Memory usage is not so impressive but it is negligible when considered when considered this as a solution to rare item problem Parameter setting procedure via quick parameter evaluation in frequent pattern mining for outbreak detection [23] Determining the minimum support required to determine the item set count and interesting patterns.…”

“…Zalizah Awang Long et al [23] described a design (shown in the above figure) to determine the minimum support required to determine the interesting pattern, minimum support and item set count. It is broadly divided in three stages: 1.Pre-processing An experiment involving three dataset, ZOO, TIC TAC TOE (UCI data repositories) and third one taken from Auto lab for WSARE implementation of outbreak detection was conducted.…”

Bench Marking Frequent Item set Mining Models and Algorithms: Current State of the Art

“…The Benchmarking models [23,37,38,39,40,41,42,43,48,49,50,25] that require support to find frequent items are computing the support through statistical approaches such as probability [37], sampling [23], averages [38,39], upper bounds [40], estimation [41,42], maximal constraints [49,50], pre-order links [25], bin oriented [48] and algorithmic [43].…”

“…Based on the result, to reduce the boring design the probable principles of lowest assistance are set at 30% to 50%. In this respect Zalizah Awang Long et al [23] described a design to figure out the exciting design; lowest assistance and product set depend. Variety of product set needed for maximum recognition amount is identified on the basis of great regularity depend.…”

“…Hence, an enhanced criteria known as the Categorized Pressure tree (SC tree) where organization rules can be excavated in a bottom-up style instead of bi-directional or recursive is suggested. Tree Based [22], [25], [26], [31], [33], [34], [35], [51], [52], [53], [54], [55], [56], [57] Apriori Based [23], [28], [29], [30], [32], [37], [38], [39], [40], [41], [42], [43], [47], [48], [49], [50] Candidature [23], [28], [29], [30], [32], [37], [38], [39], [40], [41], [42], [43], [47], [48], [49], [50] support constraint [22], [25], …”