Abstract Customer churn prediction is a method that companies use to anticipate loss in revenue. Some data mining classification techniques can be used to predict customer churn. However, these techniques could become less optimal when faced with imbalanced data conditions. Customer churn data has imbalanced data characteristics, so a process that can handle imbalanced data is required. There are two approaches... Customer churn prediction is a method that companies use to anticipate loss in revenue. Some data mining classification techniques can be used to predict customer churn. However, these techniques could become less optimal when faced with imbalanced data conditions. Customer churn data has imbalanced data characteristics, so a process that can handle imbalanced data is required. There are two approaches that can solve these problems, namely sampling method (distribution of training data is modified so that two classes of data can be balanced) and algorithm approach (algorithm process is modified to handle imbalanced data). This paper used the algorithm approach because the consistency of original data distribution will be kept the same as the training data. This will provide more valid data and prediction results that can better represent real conditions. In line with this, we proposed a Modified Balanced Random Forest (MBRF) algorithm as a classification technique to address imbalanced data. The MBRF process changes the process in a Balanced Random Forest by applying an undersampling strategy based on clustering techniques for each data bootstrap decision tree in the Random Forest algorithm. The proposed MBRF method yielded better performance compared to the Balanced Random Forest (BRF) and Random Forest (RF) algorithms, with a sensitivity value or true positive rate (TPR) of 88%, a specificity or true negative rate (TNR) of 94%, and the best AUC accuracy value of 91.65%. Moreover, MBRF also reduced process running time. Read More Read Less

Abstract Development inequality is a situation in which there are differences in development achievements between one region and another. There are many approaches to measure and to classify development inequality, but there is no focus on measuring inequality based on distance between data of development achievement, such as using data of Gross Regional Domestic Product (GRDP). This research aims to classify regional development inequality... Development inequality is a situation in which there are differences in development achievements between one region and another. There are many approaches to measure and to classify development inequality, but there is no focus on measuring inequality based on distance between data of development achievement, such as using data of Gross Regional Domestic Product (GRDP). This research aims to classify regional development inequality based on distance between data of development achievement using Fuzzy-Klassen approach. Analysis data sample is GRDP of West Java in 2015 and 2016. The result of this research shows that there are 89% of regencies/cities included in rapid-developing category (K1), 7% are in the category of developed but depressed (K2), while 4% of them are in the category of underdeveloped regions (K4). Read More Read Less

Abstract Based on vehicle – following model, we propose a new traffic flow dynamics model from both micro and macro aspects. The model can overcome some of the problems that exist in the existing kinetic models, such as avoiding the phenomenon that the traffic jam occurs in the vicinity of the disturbance density, thus exceeding the blockage density. Through design of examples, the authors... Based on vehicle – following model, we propose a new traffic flow dynamics model from both micro and macro aspects. The model can overcome some of the problems that exist in the existing kinetic models, such as avoiding the phenomenon that the traffic jam occurs in the vicinity of the disturbance density, thus exceeding the blockage density. Through design of examples, the authors also simulated traffic flow under different conditions, which shows the model has good numerical simulation capabilities, thus proving its theoretical and practical value. Read More Read Less