Preetham Kumar

Discovery of weighted association rules mining

Mining of association rules for basket databases, has been investigated by [1] [3] [4], [9], [12], etc. Most of these works focus on mining binary association rules, i.e, most of the association rules mining algorithms to discover frequent itemsets do not consider the quantity in which items have been purchased. This paper discusses an efficient method for discovering a weighted association rules from a large volumes of data in a single scan of the database. The data structure used here is called Weighted Tree. We found that this algorithm is more efficient than Cais Algorithm.

Bitwise dynamic itemset counting algorithm

Data mining has gained a lot of importance as well as popularity in todays world. Data mining provides a systematic approach for gathering useful information from huge amounts of data. Many algorithms are being written for this purpose. One of them is Dynamic Itemset Counting Algorithm. Only if all the subsets are frequent, an itemset is considered frequent in this algorithm. As the itemsets are counted, they are grouped together into four separate categories namely, dashed circle, dashed box, solid circle, and solid box. Here, a variation of this existing algorithm is being provided. Bitwise Dynamic Itemset Counting Algorithm aims to modify the existing algorithm such that its time complexity reduces. In todays world, it is very important not only to collect information from raw data but also to do it fast. Time required for running any algorithm on a collection of data directly impacts the usefulness of that algorithm. Hence, reduction of the time complexity of an existing data mining algorithm such as Dynamic Itemset Counting Algorithm shall be useful. In the existing algorithm, all transactions are checked during every pass for detecting the frequency of the different itemsets. The modified algorithm attempts to suggest a more efficient way to achieve the same results. It also aims at reducing the number of comparisons and the required number of scans. Bitwise Dynamic Itemset Counting Algorithm uses bitwise mapping of all transactions corresponding to each distinct item and the possibility check.

Mining single pass weighted pattern tree

Weighted tree mining has become an important research topic in Data mining. There are several algorithms for mining Frequent Pattern trees. FP growth algorithm using FP tree has been considered for frequent pattern mining because of its enormous performance and development compared to the candidate generation model of Apriori. The purpose of our work is to provide a tree structure for incremental and interactive weighted pattern mining by only one database scan. It is applied to existing Compact pattern (CP) tree. CP tree dynamically achieves frequency-descending prefix tree structure with a single-pass by applying tree restructuring technique and considerably reducing the mining time. It is competent of using prior tree structures and acquires mining outcomes to decrease the computation by incredible amount. Performance analysis show that our tree structure is very efficient for incremental and interactive weighted pattern mining.

Attribute -TID method for discovering sequence of attributes

The abstraction based algorithms read databases in sequential order and then construct abstraction of the database in memory. Given any database with n attributes, it is possible to read the same in n! ways. These different n! ways lead to abstractions of different sizes. In this paper, for a given a set of transactions D, we find the sequence or order of the attributes in which the database is read, a representation which is compact than PC-tree, can be obtained in the memory.

Parallel Method for Discovering Frequent Itemsets Using Weighted Tree Approach

Every element of the transaction in a transaction database may contain the components such as item number, quantity, cost of the item bought and some other relevant information of the customer. Most of the association rules mining algorithms to discover frequent itemsets do not consider the components such as quantity, cost etc. In a large database it is possible that even if the itemset appears in a very few transactions, it may be purchased in a large quantity. Further, this may lead to very high profit. Therefore these components are the most important information and without which it may cause the lose of information. This motivated us to propose a parallel algorithm to discover all frequent itemsets based on the quantity of the item bought in a single scan of the database. This method achieves its efficiency by applying two new ideas. Firstly, transaction database is converted into an abstraction called Weighted Tree that prevents multiple scanning of the database during the mining phase. This data structure is replicated among the parallel nodes. Secondly, for each frequent item assigned to a parallel node, an item tree is constructed and frequent itemsets are mined from this tree based on weighted minimum support.

Identification and red blood cell classification using computer aided system to diagnose blood disorders

Red blood cell count plays a vital role in identifying the overall health of the patient. Mature Red blood cells undergo morphological changes when blood disorder exists. Automated and Manual techniques exist in the market to count the number of RBCs(Red blood cells). Manual counting involves the use of Hemocytometer to count the blood cells. The conventional method of placing the smear under a microscope and counting the cells manually leads to erroneous results and medical laboratory technicians are put under stress. Automated counters fail to identify abnormal cells. A computer aided system will help to attain precise results in less amount of time. This research work proposes an image processing technique to separate the Red blood cell from other components of blood. It aims to examine and process the blood smear image, in order to support the classification of Red blood cells into 11 categories. K-Medoids algorithm which is robust to external noise is used to extract the WBCs from the image. The granulometric analysis is used to separate the Red blood cells from White blood cells. Feature extraction is done to obtain the significant features that help in classification. The classification results help in diagnosing the diseases like Sickle Cell Anemia, Hereditary Spherocytosis, Normochromic Anemia, Iron Deficiency Anemia, Megaloblastic Anemia and Hypochromic Anemia within few seconds.

I2Apriori: An improved apriori algorithm based on infrequent count

In this paper, the proposed method reduces CPU computation time by reducing transaction scan. The Concept infrequent count is based on minimum threshold support and 2-way searching to reduce execution time during scanning of transaction is introduced in proposed method. There exist several data mining algorithms for finding association rules but one of the candidate generation algorithms named Apriori algorithm is considered for the proposed work.

Data Mapping in Intelligent Form Using Random Hierarchical Bit Format Enhancing the Security in Data Retrieval

The paper highlights a unique and secure way of retrieving data, where user can never identify the path followed by the system to reach the goal. The whole collection of data is represented in Boolean form which is stored in a matrix. This matrix can be used for quick retrieval of data as all the functions and properties of mathematical logic can be applied. This also helps a great deal in saving memory as the data is stored in Boolean form. Boolean answers to the auto generated questions asked by computer produces a path. The main advantage of this concept is, the data is highly secured as only the person who has the key matrix can understand the meaning of matrix. The data is normalized so the efficiency in the data retrieving is also increased. The computer puts an auto generated random set of questions each time when the data has to be retrieved, which results in formation of a hierarchical tree. Thus any malicious attempt to use the previously used key won’t open the same lock and the attempt maker can be traced.