Since the problem was stated in 2004 [4] To date, there have been many mining algorithms highly useful collection was developed to improve the efficiency of extraction: UMining (2004) [4], UMining-H ( 2006) [5], Two-Phase (2005) [6], IHUP (2009) [7], TWU-Mining (2009) [8], UP-Growth (2010) [9], DTWU-Mining (2011) [10], Efim (2015) [11] and a number of other development of useful set high, as mining typical episode closed with Chui (2011) [12], AprioriCH, AprioriHC-D (2015) [13 ]; Top-k mining HUI has TKU (2012) [14], TKO (2016) [15]; HUI mining on data stream with punched-Mine (2008) [16], GUIDE (2012) [17], or HUI mining data on uncertainty [18].
All of the above algorithm assumes that profit of the items does not change in all transactions (such as Table 1a), ie if the item {b} be purchased in the transaction T1 for $ 2 ($ instance) {b} also the price of 2 ($) when purchased in the transaction T2, T4 and T5. In fact, the profit of the item may vary depending on the particular transaction due to changes in prices over time, due to promotions, etc. An example of the change is for the profit of the items in Table 2.
đang được dịch, vui lòng đợi..
