Bac Le

A new method for mining Frequent Weighted Itemsets based on WIT-trees

The mining frequent itemsets plays an important role in the mining of association rules. Frequent itemsets are typically mined from binary databases where each item in a transaction may have a different significance. Mining Frequent Weighted Itemsets (FWI) from weighted items transaction databases addresses this issue. This paper therefore proposes algorithms for the fast mining of FWI from weighted item transaction databases. Firstly, an algorithm for directly mining FWI using WIT-trees is presented. After that, some theorems are developed concerning the fast mining of FWI. Based on these theorems, an advanced algorithm for mining FWI is proposed. Finally, a Diffset strategy for the efficient computation of the weighted support for itemsets is described, and an algorithm for mining FWI using Diffsets presented. A complete evaluation of the proposed algorithms is also presented.

DBV-Miner: A Dynamic Bit-Vector approach for fast mining frequent closed itemsets

Frequent closed itemsets (FCI) play an important role in pruning redundant rules fast. Therefore, a lot of algorithms for mining FCI have been developed. Algorithms based on vertical data formats have some advantages in that they require scan databases once and compute the support of itemsets fast. Recent years, BitTable (Dong & Han, 2007) and IndexBitTable (Song, Yang, & Xu, 2008) approaches have been applied for mining frequent itemsets and results are significant. However, they always use a fixed size of Bit-Vector for each item (equal to number of transactions in a database). It leads to consume more memory for storage Bit-Vectors and the time for computing the intersection among Bit-Vectors. Besides, they only apply for mining frequent itemsets, algorithm for mining FCI based on BitTable is not proposed. This paper introduces a new method for mining FCI from transaction databases. Firstly, Dynamic Bit-Vector (DBV) approach will be presented and algorithms for fast computing the intersection between two DBVs are also proposed. Lookup table is used for fast computing the support (number of bits 1 in a DBV) of itemsets. Next, subsumption concept for memory and computing time saving will be discussed. Finally, an algorithm based on DBV and subsumption concept for mining frequent closed itemsets fast is proposed. We compare our method with CHARM, and recognize that the proposed algorithm is more efficient than CHARM in both the mining time and the memory usage.

An efficient and effective algorithm for mining top-rank-k frequent patterns

Using N-list structure for mining top-rank-k frequent patterns effectively.Subsume concept was also used to speed up the runtime of the mining process.The experiment was conducted to show the effectiveness of the proposed algorithm. Frequent pattern mining generates a lot of candidates, which requires a lot of memory usage and mining time. In real applications, a small number of frequent patterns are used. Therefore, the mining of top-rank-k frequent patterns, which limits the number of mined frequent patterns by ranking them in frequency, has received increasing interest. This paper proposes the iNTK algorithm, which is an improved version of the NTK algorithm, for mining top-rank-k frequent patterns. This algorithm employs an N-list structure to represent patterns. The subsume concept is used to speed up the process of mining top-rank-k patterns. The experiments are conducted to evaluate iNTK and NTK in terms of mining time and memory usage for eight datasets. The experimental results show that iNTK is more efficient and faster than NTK.

Interestingness measures for association rules: Combination between lattice and hash tables

There are many methods which have been developed for improving the time of mining frequent itemsets. However, the time for generating association rules were not put in deep research. In reality, if a database contains many frequent itemsets (from thousands up to millions), the time for generating association rules is more longer than the time for mining frequent itemsets. In this paper, we present a combination between lattice and hash tables for mining association rules with different interestingness measures. Our method includes two phases: (1) building frequent itemsets lattice and (2) generating interestingness association rules by combining between lattice and hash tables. To compute the measure value of a rule fast, we use the lattice to get the support of the left hand side and use hash tables to get the support of the right hand side. Experimental results show that the mining time of our method is more effective than the method that of directly mining from frequent itemsets uses hash tables only.

Mining traditional association rules using frequent itemsets lattice

There are many methods which have been developed for improvement of time in mining frequent itemsets. However, the methods which deal with the time of mining association rules were not put in deep research. In reality, in case of database which contains many frequent itemsets (from ten thousands up to millions), the time of mining association rules is much larger than that needed for mining frequent itemsets. In this paper, we present an application of lattice in mining traditional association rules which will reduce greatly the time for mining rules - our method includes two phases: (1) building frequent itemsets lattice and (2) mining association rules from lattice. We based on the parent-child relationships in lattice to fast discover the association rules. The experiments show that the mining rules from lattice is more effective than the direct mining from frequent itemsets using hash table.

A Novel Algorithm for Mining High Utility Itemsets

The utility based itemset mining approach has been discussed widely in recent years. There are many algorithms mining high utility itemsets by pruning candidates based on estimated utility values, and based on transaction-weighted utilization values. These algorithms aim to reduce search space. Besides, candidate pruning based on transaction-weighted utilization value is better than other strategies. In this paper, we propose TWU-Mining, a novel algorithm based-on WIT-tree for improving the cost of time and search space. Experiments show that the proposed algorithm is more effective on the testing databases.

Incrementally building frequent closed itemset lattice

A concept lattice is an ordered structure between concepts. It is particularly effective in mining association rules. However, a concept lattice is not efficient for large databases because the lattice size increases with the number of transactions. Finding an efficient strategy for dynamically updating the lattice is an important issue for real-world applications, where new transactions are constantly inserted into databases. To build an efficient storage structure for mining association rules, this study proposes a method for building the initial frequent closed itemset lattice from the original database. The lattice is updated when new transactions are inserted. The number of database rescans over the entire database is reduced in the maintenance process. The proposed algorithm is compared with building a lattice in batch mode to demonstrate the effectiveness of the proposed algorithm.

Integrated Uncertainty in Knowledge Modelling and Decision Making

• Uncertainty formalisms: Bayesian probability, Dempster-Shafer theory, imprecise probability, random sets, rough sets, fuzzy sets & interval-based models. • Modelling uncertainty & inconsistency in big data • Learning and reasoning with uncertainty • Information fusion & knowledge integration in uncertain environments • Decision making under various types of uncertainty • Aggregation operators • Copulas for dependence modelling • Granular and soft computing • Computational intelligence Application

An effective approach for maintenance of pre-large-based frequent-itemset lattice in incremental mining

Incremental mining has attracted the attention of many researchers due to its usefulness in online applications. Many algorithms have thus been proposed for incrementally mining frequent itemsets. Maintaining a frequent-itemset lattice (FIL) is difficult for databases with large numbers of frequent itemsets, especially huge databases, due to the storage of links of nodes in the lattice. However, generating association rules from a FIL has been shown to be more effective than traditional methods such as directly generating rules from frequent itemsets or frequent closed itemsets. Therefore, when the number of frequent itemsets is not huge (i.e., they can be stored in the lattice without excessive memory overhead), the lattice-based approach outperforms approaches which mine association rules from frequent itemsets/frequent closed itemsets. However, incremental algorithms for building FILs have not yet been proposed. This paper proposes an effective approach for the maintenance of a FIL based on the pre-large concept in incremental mining. The building process of a FIL is first improved using two proposed theorems regarding the paternity relation between two nodes in the lattice. An effective approach for maintaining a FIL with dynamically inserted data is then proposed based on the pre-large and the diffset concepts. The experimental results show that the proposed approach outperforms the batch approach for building a FIL in terms of execution time.

Multi-scale Sparse Representation for Robust Face Recognition

Recently the Sparse Representation-based Classification (SRC) has been successfully used in face recognition. In SRC, a test image is coded by a linear combination of the training dictionary. In this paper, we propose a model extends from SRC named Multi-scale SRC (MSRC). The MSRC build the multi-scale dictionary for the training. A test image is then coded using this multi-scale dictionary. In addition, a voting scheme is applied which not only helps improving the recognition rate significantly, but also makes the algorithm more robust with occlusion. Experiments on representative face databases demonstrate that the MSRC is much more effective than the SRC.