A novel distance-based classifier built on pattern ranking

Pattern Recognition Letters, 2010

In this paper we present InstanceRank, a ranking algorithm that reflects the relevance of the instances within a dataset. InstanceRank applies a similar solution to that used by PageRank, the web pages ranking algorithm in the Google search engine. We also present ISR, an instance selection technique that uses InstanceRank. This algorithm chooses the most representative instances from a learning database. Experiments show that ISR algorithm, with InstanceRank as ranking criteria, obtains similar results in accuracy to other instance reduction techniques, noticeably reducing the size of the instance set.

ArXiv, 2017

The last decade has seen a revolution in the theory and application of machine learning and pattern recognition. Through these advancements, variable ranking has emerged as an active and growing research area and it is now beginning to be applied to many new problems. The rationale behind this fact is that many pattern recognition problems are by nature ranking problems. The main objective of a ranking algorithm is to sort objects according to some criteria, so that, the most relevant items will appear early in the produced result list. Ranking methods can be analyzed from two different methodological perspectives: ranking to learn and learning to rank. The former aims at studying methods and techniques to sort objects for improving the accuracy of a machine learning model. Enhancing a model performance can be challenging at times. For example, in pattern classification tasks, different data representations can complicate and hide the different explanatory factors of variation behin...

International Journal of Computer Applications, 2012

In the field of computer science, data mining is the process that attempts to discover patterns in large data sets. However it deals mostly with the relationship between two or more item objects. For example A to B, where 'A' and 'B' are the item objects. But in the real life scenario not only the relationship between item objects is important, but the relationship of their frequency of occurrence is also the matter of a prime concern. The instances of two or more data items also may be correlated with each other. For example the relation between A and 2B.Where 'A' and 'B' are the data items and '2B' represents two instances of the B type of data items. This paper provides a new approach to find the occurrence dependent data patterns by conventional approaches and also compare the some inter related concepts.

Computer Science and Information Systems, 2020

Huge amounts of data are being collected and analyzed nowadays. By using the popular rule-learning algorithms, the number of rule discovered on those ?big? datasets can easily exceed thousands. To produce compact, understandable and accurate classifiers, such rules have to be grouped and pruned, so that only a reasonable number of them are presented to the end user for inspection and further analysis. In this paper, we propose new methods that are able to reduce the number of class association rules produced by ?classical? class association rule classifiers, while maintaining an accurate classification model that is comparable to the ones generated by state-of-the-art classification algorithms. More precisely, we propose new associative classifiers, called DC, DDC and CDC, that use distance-based agglomerative hierarchical clustering as a post-processing step to reduce the number of its rules, and in the rule-selection step, we use different strategies (based on database coverage an...

International Journal of …, 2012

Frequent pattern mining has been a focused theme in data mining research for over a decade. Abundant literature has been dedicated to this research and tremendous progress has been made, ranging from efficient and scalable algorithms for frequent itemset mining in transaction databases to numerous research frontiers, such as sequential pattern mining, structured pattern mining, correlation mining, associative classification, and frequent pattern-based clustering, as well as their broad applications. In this paper, we develop a new technique for more efficient pattern mining. Our method find frequent 1-itemset and then uses the heap tree sorting we are generating frequent patterns, so that many. We present efficient techniques to implement the new approach.

Fifth IEEE International Conference on Data Mining (ICDM'05), 2005

The discovery of frequent patterns is a famous problem in data mining. While plenty of algorithms have been proposed during the last decade, only a few contributions have tried to understand the influence of datasets on the algorithms behavior. Being able to explain why certain algorithms are likely to perform very well or very poorly on some datasets is still an open question. In this setting, we describe a thorough experimental study of datasets with respect to frequent itemsets. We study the distribution of frequent itemsets with respect to itemsets size together with the distribution of three concise representations: frequent closed, frequent free and frequent essential itemsets. For each of them, we also study the distribution of their positive and negative borders whenever possible. From this analysis, we exhibit a new characterization of datasets and some invariants allowing to better predict the behavior of well known algorithms. The main perspective of this work is to devise adaptive algorithms with respect to dataset characteristics.

Lecture Notes in Computer Science, 2010

Emerging patterns are itemsets whose frequencies change sharply from one class to the other. PCL is an example of efficient classification algorithms that leverage the prediction power of emerging patterns. It first selects the top-K emerging patterns of each class that match a testing instance, and then uses these selected patterns to decide the class label of the testing instance. We study the impact of the parameter K on the accuracy of PCL. We have observed that in many cases, the value of K is critical to the performance of PCL. This motivates us to develop an algorithm to find the best value of K for PCL. Our results show that finding the best K can improve the accuracy of PCL greatly, and employing incremental frequent itemset maintenance techniques reduces the running time of our algorithm significantly.

Lecture Notes in Computer Science, 2000

Emerging patterns (EPs), namely itemsets whose supports change significantly from one class to another, capture discriminating features that sharply contrast instances between the classes. Recently, EP-based classifiers have been proposed, which first mine as many EPs as possible (called eager-learning) from the training data and then aggregate the discriminating power of the mined EPs for classifying new instances. We propose here a new, instance-based classifier using EPs, called DeEPs, to achieve much better accuracy and efficiency than the previously proposed EP-based classifiers. High accuracy is achieved because the instance-based approach enables DeEPs to pinpoint all EPs relevant to a test instance, some of which are missed by the eager-learning approaches. High efficiency is obtained using a series of data reduction and concise data-representation techniques. Experiments show that DeEPs' decision time is linearly scalable over the number of training instances and nearly linearly over the number of attributes. Experiments on 40 datasets also show that DeEPs is superior to other classifiers on accuracy. * * * * Since EPs usually have very low supports, they are not suitable to be used individually for classification. We will use the compact summation method to aggregate the discriminating power contributed by all selected EPs to form classification scores.

This paper introduces a new classification algorithm of the instance-based learning type. Training records are converted into patterns associated with a known class label, and stored permanently into a trie 1 -like tree structure along with other helpful information. Classifying new records is done selecting from the trie two best patterns as solutions hypotheses. Best pattern selection is done using standard distance metrics, a strength function and an exclusive values concept. Classification tests done on several data files have shown very accurate results.

Nowadays there is vast amount of data being collected and stored in databases and without automatic methods for extracting this information it is practically impossible to mine for them. In Data Mining, the Classification processes perchance the most recognizable and most popular concepts. Actually Classification maps data into predefined groups or classes. Classification normally uses prediction rules to express knowledge. This Prediction rules are expressed in the form of IF-THEN rules. Sometimes this classification referred to as supervised learning by reason of the classes is determined before examining the data. Classification consists of predicting a certain outcome based on a given input. In order to predict the outcome, the algorithm processes a training set containing a set of attributes and the respective outcome, usually called goal or prediction attribute. Classification problems handled by using some known type of classification algorithms such that Statistical-Based Algorithms, Distance Based Algorithms etc,. All approaches to performing classification assume some knowledge of the data. This paper will focus on Distance Based Algorithms in classification. In Distance Based Algorithms each item is mapped to the same class may be thought of as more similar to the other items in that class than items found in other classes. Therefore, similarity or distance measures may be used to identify the "alikeness" of different items in the database. Using a similarity measure for classification where the classes are predefined is somewhat simpler than using a similarity measure for clustering where the classes are not known in advance. So the classification problem then becomes one of determining similarity not among all tuples in the database but between each tuple and the query