In competitive telecommunication environment, it is imperative to maintain an effective customer retention strategy even while mobile service operators attracting new customers. Not only acquiring new customers costly process, but successful customer retention helps build brand loyalty and good business reputation. Motivated by real mobile service operator data set, we designed and proposed a solution to employ survival analysis technique that estimates customers’ survivals and hazards. We aim to examine the impact of: campaign, tariff, tenure, age, auto-payment on survival times and hazards. After hazard ratios and survival experiences determined for each predictor, results enable mobile service operator to target the right customers to incentivize so that they can stay with their current operator. Proactive actions triggered by the results of the survival model is key to customer retention.

Traffic accidents, which continue to increase from year to year in Turkey and in the world, have become a huge problem that can result in serious traumas, injuries, and deaths, as well as their material and moral consequences. Many studies have been carried out in the world and in Turkey to reduce the number of traffic accidents, but these studies have not been very effective in reducing accidents. In this study, 3105 fatal or injured traffic accidents between 2010–2017 in Sarıyer district of Istanbul, Turkey’s largest city in terms of population, were discussed. We analyzed the statistical information on the subject in detail within the framework of geographic information systems. It has been tried to determine the sections where traffic accidents are concentrated in this region with studies based on spatial methods. Thematic accident map was created according to the accident types. In this context, the advantages and disadvantages of these methods were compared using Point Density, Kernel Density, Getis Ord Gi*, and Anselin Local Moran’s I (LISA) Spatial Autocorrelation. In addition, in order to observe the change in accidents, thematic accident and Kernel Density maps were created separately according to accident occurrence types in the beginning and last year. From this point of view, the changes that occurred in the accidents were interpreted. The current study determined that the most accidents were on some streets and these streets divided into regions in a plan. The cases were examined with statistical analyses according to accident types and using the Kernel Density method. In addition, it has been observed that Kernel Density method gives better visual results than other spatial methods. In this study, spatial analysis and statistical analysis methods were used to evaluate traffic accidents more realistically. The day of the week effect and month of the year effect on traffic accidents was investigated for the first time. In addition, it is proposed to bring a new approach to the prevention of traffic accidents by using hotspot, accident type, and day of the week effect.

In this study, a hybrid model combining discrete wavelet transforms (WTs) and artificial neural networks (ANNs) is used to estimate the monthly streamflow. The WT-ANN hybrid model was developed using the Daubechies main wavelet to predict the streamflow for three gauging stations on the Çoruh river basin one month in advance, with different combinations of air temperature, precipitation, and streamflow variables, and their wavelet transformations. Four different hybrid WT-ANN models were generated and compared with four different conventional ANN models. The dataset was chronologically divided into training, validation, and testing data. The results indicated that the WT-ANN hybrid models performed better than the traditional ANN models for all three stations. Furthermore, the chronologically divided dataset was used to examine the effects of changes in hydrological data over time on model performance. In conclusion, model performances in the training period deteriorated during the validation and testing periods due to structural changes in the hydrological data.