Somayeh Alizadeh

Comparing simulated annealing and genetic algorithm in learning FCM

Fuzzy Cognitive Map (FCM) is a directed graph, which shows the relations between essential components in complex systems. It is a very convenient, simple, and powerful tool, which is used in numerous areas of application. Experts who are familiar with the system components and their relations can generate a related FCM. There is a big gap when human experts cannot produce FCM or even there is no expert to produce the related FCM. Therefore, a new mechanism must be used to bridge this gap. In this paper, a novel learning method is proposed to construct FCM by using some metaheuristic methods such as genetic algorithm (GA) and simulated annealing (SA). The proposed method not only is able to construct FCM graph topology but also is able to extract the weight of the edges from input historical data. The efficiency of the proposed method is shown via comparison of its results of some numerical examples with those of some other methods.

Estimating customer lifetime value based on RFM analysis of customer purchase behavior: Case study

Abstract Since the increased importance is placed on customer equity in today’s business environment, many firms are focusing on the notion of customer loyalty and profitability to increasing market share. Building successful customer relationship management (CRM), a firm starts from identifying customers’ true value and loyalty since customer value can provide basic information to deploy more targeted and personalized marketing. In this paper, customer lifetime value (CLV) is used to customer segmentation of a health and beauty company. Two approaches are used: in the first approach, RFM (Recency, Frequency, and Monetary) marketing analysis method is used in order to segmentation of customers and in the second approach, the proposed extended RFM analysis method with one additional parameter—called Count Item—is used. Comparing results of these approaches, shows that adding count Item as a new parameter to RFM method makes no difference to clustering result, so CLV is calculated based on weighted RFM method for each segment. The results of calculated CLV for different segments can be used to explain marketing and sales strategies by the company.

Learning Fuzzy Cognitive Maps using Imperialist Competitive Algorithm

Abstract In this paper, a new automated Fuzzy Cognitive Maps (FCMs) learning algorithm is developed to generate FCMs from historical data. Automated FCM learning algorithms are used to model and analyze systems which are very complex and cannot be handled by experts’ knowledge. The algorithm developed in this paper is based on the Imperialist Competitive Algorithm for global optimization and is called the Imperialist Competitive Learning Algorithm (ICLA). The ICLA divides the search space into several sections. It extracts the best knowledge from each section and follows a procedure to avoid local optima alongside rapid learning. Experiments have been conducted to compare the ICLA with other well-known FCM learning algorithms. The results show that in most cases, the ICLA performs better for learning FCMs in terms of solution accuracy and execution time. The testing results show clearly that the ICLA is a robust, fast and accurate FCM learning algorithm.

Data-mining application for country segmentation based on the RFM model

For effective Customer Relationship Management (CRM), it is important to gather information on customer value. Segmentation is the method of knowing the customers and partitioning a population of customers into smaller groups. This paper develops a novel country segmentation methodology based on Recency (R), Frequency (F) and Monetary value (M) variables. After the variables are calculated, clustering methods (K-means and fuzzy K-means) are used to segment countries and compare the results of these methods by three different criteria. Customers are classified into four tiers: Top-active, Medium-active, New customer and Inactive. Then a customer pyramid is drawn and the customer value is calculated. Consequently, the data are used to analyse the relative profitability of each customer cluster and the proper strategy is determined for them.

Type 2 Diabetes Mellitus Screening and Risk Factors Using Decision Tree: Results of Data Mining

Objectives: The aim of this study was to examine a predictive model using features related to the diabetes type 2 risk factors. Methods: The data were obtained from a database in a diabetes control system in Tabriz, Iran. The data included all people referred for diabetes screening between 2009 and 2011. The features considered as “Inputs” were: age, sex, systolic and diastolic blood pressure, family history of diabetes, and body mass index (BMI). Moreover, we used diagnosis as “Class”. We applied the “Decision Tree” technique and “J48” algorithm in the WEKA (3.6.10 version) software to develop the model. Results: After data preprocessing and preparation, we used 22,398 records for data mining. The model precision to identify patients was 0.717. The age factor was placed in the root node of the tree as a result of higher information gain. The ROC curve indicates the model function in identification of patients and those individuals who are healthy. The curve indicates high capability of the model, especially in identification of the healthy persons. Conclusions: We developed a model using the decision tree for screening T2DM which did not require laboratory tests for T2DM diagnosis.

ICLA imperialist competitive learning algorithm for fuzzy cognitive map: Application to water demand forecasting

In this paper, we develop a new Fuzzy Cognitive Map (FCM) learning method using the imperialistic competitive learning algorithm (ICLA). An FCM seems like a fuzzy signed directed graph with feedback, and models complex systems as a collection of concepts and causal relations between concepts. Conventional FCMs are mainly constructed by human experts who have experience in the specific problem domain. However, large problems need automated methods. We develop an automated method for FCM construction inspired by the socio-political behavior of countries as imperialists with colonies. In the real world imperialists extend their territories and change the socio attributes of their colonies. The ICLA is an evolutionary algorithm and simulates this behavior. We explain the algorithm for FCM learning and demonstrate its performance advantages through synthetic and real data of water demand. The results of the new algorithm were compared to that of a genetic algorithm, which is the most commonly used and well-known FCM learning algorithm.

Mining important association rules based on the RFMD technique

The method of association rule mining has been used by marketers for many years to extract marketing rules from purchase transactions. Marketers and managers employ these rules in order to predict customer needs for future sales. Extracting effective rules is one of the major problems of marketers. Effective rules can help them to make better marketing decisions. On the other hand, the Recency, Frequency, Monetary value and Duration (RFMD) method is one of the popular methods used in market segmentation that indicate profitable groups of customers. In this paper, a novel method is proposed that takes advantage of the RFMD method to extract effective association rules from profitable segments of purchase transactions. In other words, in the first step, raw data are classified based on the RFMD technique; and in the second step, effective association rules are extracted from sections with high RFMD values. The proposed method employs a new Maximum Frequent Itemset Extractor (MFIE) algorithm that outperforms the classic algorithm (Apriori) in extracting frequent itemsets from a large number of transactions. In addition, unlike most of the previous central methods, the proposed method is designed for extracting association rules from distributed databases.

Web Spam Detection by Learning from Small Labeled Samples

Web spamming tries to deceive search engines to rank some pages higher than they deserve. Many methods have been proposed to combat web spamming and to detect spam pages. One basic method is using classification, i.e., learning a classification model from previously labeled training data and using this model for classifying web pages to spam or nonspam. A drawback of this method is that manually labeling a large number of web pages to generate the training data can be biased, non-accurate, labor intensive and time consuming. In this paper, we are going to propose a new method to resolve this drawback by using semi-supervised learning to automatically label the training data. To do this, we incorporate Expectation-Maximization algorithm that is an efficient and an important algorithm of semi-supervised learning. Experiments are carried out on the real web spam data, which show the new method, performs very well in practice. General Terms Information Retrieval, Search Engine, Machine Learning.

Clustering Networks’ Heterogeneous Data in Defining a Comprehensive Closeness Centrality Index

One of the most important applications of network analysis is detecting community structure, or clustering. Nearly all algorithms that are used to identify these structures use information derived from the topology of these networks, such as adjacency and distance relationships, and assume that there is only one type of relation in the network. However, in reality, there are multilayer networks, with each layer representing a particular type of relationship that contains nodes with individual characteristics that may influence the behavior of networks. This paper introduces a new, efficient spectral approach for detecting the communities in multilayer networks using the concept of hybrid clustering, which integrates multiple data sources, particularly the structure of relations and individual characteristics of nodes in a network, to improve the comprehension of the network and the clustering accuracy. Furthermore, we develop a new algorithm to define the closeness centrality measure in complex networks based on a combination of two approaches: social network analysis and traditional social science approach. We evaluate the performance of our proposed method using four benchmark datasets and a real-world network: oil global trade network. The experimental results indicated that our hybrid method is sufficiently effective at clustering using the node attributes and network structure.

A new model for discovering process trees from event logs

Process mining techniques aim at extracting knowledge from event logs. One of the most important tasks in process mining is process model discovery. In discovering process models, an algorithm is designed to build a process model from a given event log. In this paper, a new model to discover process models has been proposed. A combination of Genetic Algorithm and Simulated Annealing has been used in this model. Genetic Algorithms has previously been used in this context. Previous approaches had drawbacks in fitness evaluation that misguided the algorithm. Another problem was that the quality of the candidates, in the population, was low such that it reduced the chance of finding a perfect answer. In this paper, a new fitness measure has been proposed to evaluate process models based on event logs. Moreover SA has been used to improve the quality of candidates in the population. It has been demonstrated that the proposed model outperformed in terms of rediscovering process models, compared to other approaches which are proposed in the literature, which was the result of better fitness evaluation and increased quality of individuals,. It came to conclusion that using GA and SA in combination with each other can be effective in this context.