Decision Tree Data Mining Technique Research Articles

Integration of Data Mining Techniques to PostgreSQL Database Manager System

Data mining is a technique that allows to obtain patterns or models from the gathered data. This technique is applied in all kind of environments such as in the biological field, educational and financial applications, industry, police, and political processes. Within data mining there are several techniques, among which are the induction of rules and decision trees which, according to various studies carried out, are among the most used. This research analyzes decision tree data mining techniques and induction rules to integrate several of its algorithms into PostgreSQL database management system (DBMS). Through an experiment, it was found that when the algorithms are integrated to the manager, the response times and the results obtained are higher.

An Approach for Determining Rules used to Select Viable Junction Design Alternatives Based on Multiple Objectives

Transport planners and engineers frequently face the challenge to determine the best design for a specific junction. Many road design manuals provide guidelines for the design and evaluation of different junction alternatives, however these mostly refer to specialized software in which the performances of design alternatives can be modelled. In the first stage of the design process, such assessments of many alternatives are undesirable due to time and budget constraints. There is a need for quick design rules which need limited input data. Although some of these rules exist, their usability is limited due to inconsistencies in rules and non-transparency in combination with objectives. In this paper, we present an approach by which consistent and transparent junction design rules can be determined. The resulting rules can be used to predict a set of viable junction design alternatives for the first stage of the junction design assessment process. The predicted set is in fact the Pareto optimal set of solutions for multiple objectives, e.g. regarding operational, safety and/or environmental impact. The Pareto optimal set of solutions always contains the best solution, whatever set of weights is used for different objectives in a later stage of the assessment process, thus handling multiple objectives in a straightforward manner. The rules are derived from a dataset by using decision tree data mining techniques. For this, a large dataset is first generated, using performance models, with Pareto optimal sets of junction design alternatives for a large amount of, randomly generated, traffic volumes. The approach is applied and evaluated on cases for two different countries. Results show that for over 90% of the situations the Pareto optimal set can be predicted by the new rules, whereas existing rules hardly reach 33%. The new rules provide junction design alternatives with a better performance.

An efficient incremental learning mechanism for tracking concept drift in spam filtering.

This research manages in-depth analysis on the knowledge about spams and expects to propose an efficient spam filtering method with the ability of adapting to the dynamic environment. We focus on the analysis of email’s header and apply decision tree data mining technique to look for the association rules about spams. Then, we propose an efficient systematic filtering method based on these association rules. Our systematic method has the following major advantages: (1) Checking only the header sections of emails, which is different from those spam filtering methods at present that have to analyze fully the email’s content. Meanwhile, the email filtering accuracy is expected to be enhanced. (2) Regarding the solution to the problem of concept drift, we propose a window-based technique to estimate for the condition of concept drift for each unknown email, which will help our filtering method in recognizing the occurrence of spam. (3) We propose an incremental learning mechanism for our filtering method to strengthen the ability of adapting to the dynamic environment.

An intelligent three‐phase spam filtering method based on decision tree data mining

AbstractIn this paper, we proposed an efficient spam filtering method based on decision tree data mining technique, analyzed the association rules about spams, and applied these rules to develop a systematized spam filtering method. Our method possessed the following three major superiorities: (i) checking only an e‐mail's header section to avoid the low‐operating efficiency in scanning an e‐mail's content. Moreover, the accuracy of filtering was enhanced simultaneously. (ii) In order that the probable misjudgment in identifying an unknown e‐mail could be “reversed”, we had constructed a reversing mechanism to help the classification of unknown e‐mails. Thus, the overall accuracy of our filtering method will be increased. (iii) Our method was equipped with a re‐learning mechanism, which utilized the supervised machine learning method to collect and analyze each misjudged e‐mail. Therefore, the revision information learned from the analysis of misjudged e‐mails incrementally gave feedback to our method, and its ability of identifying spams would be improved. Copyright © 2016 John Wiley & Sons, Ltd.

Information Decay in Building Predictive Models Using Temporal Data

Most predictive data mining methods are based on the assumption that the historical data involved in building and verifying a model are the best estimator of what will happen in the future. However, the relevance of past to the future depends on the application domain in a specific timeframe. In this paper, we present a method of predicting the future from temporal data using information decay technique that estimate the relevance of the past to the future with the help of the knowledge of the information lifetime of the temporal data. We show how to use a decision tree data mining technique to build an information-decay-based predictive model by introducing an information decay method in impurity measure functions. The relevance of the concepts is proven by comparing two decision tree-based classifiers: one built using information decay over a specific timeframe and the other built setting aside the time factor.

Data-Driven Decision Tree Classification for Product Portfolio Design Optimization

The formulation of a product portfolio requires extensive knowledge about the product market space and also the technical limitations of a company’s engineering design and manufacturing processes. A design methodology is presented that significantly enhances the product portfolio design process by eliminating the need for an exhaustive search of all possible product concepts. This is achieved through a decision tree data mining technique that generates a set of product concepts that are subsequently validated in the engineering design using multilevel optimization techniques. The final optimal product portfolio evaluates products based on the following three criteria: (1) it must satisfy customer price and performance expectations (based on the predictive model) defined here as the feasibility criterion; (2) the feasible set of products/variants validated at the engineering level must generate positive profit that we define as the optimality criterion; (3) the optimal set of products/variants should be a manageable size as defined by the enterprise decision makers and should therefore not exceed the product portfolio limit. The strength of our work is to reveal the tremendous savings in time and resources that exist when decision tree data mining techniques are incorporated into the product portfolio design and selection process. Using data mining tree generation techniques, a customer data set of 40,000 responses with 576 unique attribute combinations (entire set of possible product concepts) is narrowed down to 46 product concepts and then validated through the multilevel engineering design response of feasible products. A cell phone example is presented and an optimal product portfolio solution is achieved that maximizes company profit, without violating customer product performance expectations.

An efficient two-phase spam filtering method based on e-mails categorization

The e-mail's header session usually contains important attributes such as e-mail title, sender's name, sender's e-mail address, sending date, which are helpful to classification of e-mails. In this paper, we apply decision tree data mining technique to header's basic attributes to analyze the association rules of spam e-mails and propose an efficient spam filtering method to accurately identify spam and legitimate e-mails. According to the experiment of applying numerous Chinese e-mails to our spam filtering method, we obtain the following excellent datums: the Accuracy is 96.5%, the Precision is 96.67%, and the Recall is 96.3%. Thus, the method proposed in this paper can efficiently identify the spam e-mails by checking only the header sessions, which can reduce the cost for calculation.

Applying decision tree data mining for online group buying consumers' behaviour

This paper treated 'amount of invested money' and 'use times' of online group buying as the dependent variables, and applied the clustering algorithm and the decision tree data mining technique to deeply analyse the influences of independent variables on each dependent variable and observe the possible association rules existing among them. We found that the consumers' past online shopping experience would directly affect their decision-making. Noticeably, the people who tended to be willing to participate in the money-saving group buying, were not the ones with lower incomes; instead, they were the ones with more incomes.

Data Mining Methods In A Metrics-deprivedInventory Transactions Environment

Over the past decade, several data mining techniques have come into use within a variety of data intensive fields. Mining transaction data to discover interesting patterns or help forecast future rare events based upon historical records is of key interest. Among the classic problems facing any supply chain is that of determining how much of a given item to keep on the shelf of a warehouse or a store, and at what level to initiate replenishment of this supply. This problem has been well examined and extensively researched. However, applying current data mining techniques to support the determination of effective inventory stock levels and economic re-order points has yet to be explored. In this paper, Naive Bayes, Bayes Network, SVM, MLP, Logistic, and J48 decision tree data mining techniques are compared with respect to their effectiveness when tasked with predicting stock shortage conditions indicated by the reduced transaction history of a bench stock inventory. These comparisons have been made using training and test data sets drawn from a set of 1.8 million transactions. Each method is applied to the same set of transaction data that spans three years of bench stock inventory activity. This set of transactions contains orders, receipts, stock outages, and stock shortages but it does not include issue transactions from the inventory. Since the time of issue or removal of the item from the inventory is not known, determining the expected demand becomes very difficult. The effectiveness of the data mining predictive capability is measured with respect to the accuracy of the prediction and the lead-time provided to the inventory planner for the given prediction. The results show similar performance among most of the data mining algorithms using “actual event” prediction evaluation criteria.

Data mining with genetic algorithms on binary trees

This paper focuses on the automatic interaction detection (AID)-technique, which belongs to the class of decision tree data mining techniques. The AID-technique explains the variance of a dependent variable through an exhaustive and repeated search of all possible relations between the (binary) predictor variables and the dependent variable. This search results in a tree in which non-terminal nodes represent the binary predictor variables, edges represent the possible values of these predictor variables and terminal nodes or leafs correspond to classes of subjects. Despite of being self-evident, the AID-technique has its weaknesses. To overcome these drawbacks a technique is developed that uses a genetic algorithm to find a set of diverse classification trees, all having a large explanatory power. From this set of trees, the data analyst is able to choose the tree that fulfils his requirements and does not suffer from the weaknesses of the AID-technique. The technique developed in this paper uses some specialised genetic operators that are devised to preserve the structure of the trees and to preserve high fitness from being destroyed. An implementation of the algorithm exists and is freely available. Some experiments were performed which show that the algorithm uses an intensification stage to find high-fitness trees. After that, a diversification stage recombines high-fitness building blocks to find a set of diverse solutions.