Mirzania, Ehsan; Achite, Mohammed; Elshaboury, Nehal; Katipoğlu, Okan Mert; Saroughi, Mohsen

In this article Mohsen Saroughi should have been denoted as a second corresponding author. The original article has been corrected.

Mirzania, Ehsan; Achite, Mohammed; Elshaboury, Nehal; Katipoğlu, Okan Mert; Saroughi, Mohsen

Predicting the groundwater level (GWL) is essential in water resource management and irrigation planning in arid and semi-arid areas. In this study, an artificial neural network (ANN) was combined with newly developed wild horse optimizer (WHO) and egret swarm optimization algorithm (ESOA) techniques to predict a one month lead-time GWL in the Tabriz plain of Iran. For the prediction of the GWL, the number of months and years, the one month lag of average temperature, evaporation, precipitation, and GWL were used as inputs. Model performances were compared using root mean square error (RMSE), Nash–Sutcliffe efficiency (NSE), coefficient of determination (R2), and relative strength ratio (RSR) statistical indicators and scatter diagrams, time series graph, violin graph, and Taylor diagram. As a result of the analysis, the most successful estimation results were obtained with the input combinations of year, month, average temperature, evaporation, precipitation, and GWL (t − 1) for the prediction of the one month lead-time GWL. According to the results of evaluation indicators in the testing phase, ANN with (R2 = 0.871, RMSE = 0.306 (m), NSE = 0.832, and RSR = 0.410), WHO–ANN (R2 = 0.932, RMSE = 0.200 (m), NSE = 0.929, and RSR = 0.267), and ESOA–ANN (R2 = 0.952, RMSE = 0.164 (m), NSE = 0.951, and RSR = 0.220). In addition, it was revealed that the ESOA–ANN hybrid model showed higher prediction success than the WHO–ANN and standalone ANN models. The study outputs contribute to decision-makers and planners for controlling land subsidence, assessing GWL and aquifer compaction, irrigation planning, and effective management of water resources.

Akgül, İsmail

Automatic age and gender estimation provides an important information to analyze real-world applications such as human–machine interaction, system access, activity recognition, and consumer profile detection. While it is easy to estimate a person’s gender from human facial images, estimating their age is difficult. In such previous challenging studies, traditional convolutional neural network (CNN) methods have been used for age and gender estimation. With the development of deep convolutional neural network (DCNN) architectures, more successful results have been obtained than traditional CNN methods. In this study, two state-of-the-art DCNN models have been developed in the field of artificial intelligence (AI) to make age and gender estimation on an imbalanced dataset of human face images. Firstly, a new model called fast description network (FINet) was developed, which has a parametrically changeable structure. Secondly, the number of parameters has been reduced by using the layer reduction approach in InceptionV3 and NASNetLarge DCNN model structures, and a second model named inception Nasnet fast identify network (INFINet) was developed by concatenating these models and the FINet model as a triple. FINet and INFINet models developed for age and gender estimation were compared with many other state-of-the-art DCNN models in AI. The most successful accuracy results in terms of both age and gender were obtained with the INFINet model (age: 61.22%, gender: 80.95% in the FG-NET dataset, age: 72.00%, gender: 90.50% in the UTKFace dataset). The results obtained in age and gender estimation with the INFINet model are much more effective than other recent state-of-the-art works. In addition, the FINet model, which has a much smaller number of parameters than the compared models, showed a classification performance that can compete with state-of-the-art methods for age and gender estimation.

Oksuz, Mehmet Kursat; Buyukozkan, Kadir; Bal, Alperen; Satoglu, Sule Itir

The capacitated p-median problem is a well-known location-allocation problem that is NP-hard. We proposed an advanced Genetic Algorithm (GA) integrated with an Initial Solution Procedure for this problem to solve the medium and large-size instances. A 33 Full Factorial Design was performed where three levels were selected for the probability of mutation, population size, and the number of iterations. Parameter tuning was performed to reach better performance at each instance. MANOVA and Post-Hoc tests were performed to identify significant parameter levels, considering both computational time and optimality gap percentage. Real data of Lorena and Senne (2003) and the data set presented by Stefanello et al. (2015) were used to test the proposed algorithm, and the results were compared with those of the other heuristics existing in the literature. The proposed GA was able to reach the optimal solution for some of the instances in contrast to other metaheuristics and the Mat-heuristic, and it reached a solution better than the best known for the largest instance and found near-optimal solutions for the other cases. The results show that the proposed GA has the potential to enhance the solutions for large-scale instances. Besides, it was also shown that the parameter tuning process might improve the solution quality in terms of the objective function and the CPU time of the proposed GA, but the magnitude of improvement may vary among different instances.

Oksuz, Mehmet Kursat; Buyukozkan, Kadir; Bal, Alperen; Satoglu, Sule Itir

Unfortunately, the affiliation details of the third author, Alperen Bal has been published with incorrect information. The corrected affiliation information are given below: College of Engineering and Technology, American University of the Middle East, Egaila 54200, Kuwait. The original article has been corrected accordingly.