Deep Neural Networks (DNNs) are widely regarded as the most effective learning tool for dealing with large datasets, and they have been successfully used in thousands of applications in a variety of fields. Based on these large datasets, they are trained to learn the relationships between various variables. The adaptive moment estimation (Adam) algorithm, a highly efficient adaptive optimization algorithm, is widely used as a learning algorithm in various fields for training DNN models. However, it needs to improve its generalization performance, especially when training with large-scale datasets. Therefore, in this paper, we propose HN Adam, a modified version of the Adam Algorithm, to improve its accuracy and convergence speed. The HN_Adam algorithm is modified by automatically adjusting the step size of the parameter updates over the training epochs. This automatic adjustment is based on the norm value of the parameter update formula according to the gradient values obtained during the training epochs. Furthermore, a hybrid mechanism was created by combining the standard Adam algorithm and the AMSGrad algorithm. As a result of these changes, the HN_Adam algorithm, like the stochastic gradient descent (SGD) algorithm, has good generalization performance and achieves fast convergence like other adaptive algorithms. To test the proposed HN_Adam algorithm performance, it is evaluated to train a deep convolutional neural network (CNN) model that classifies images using two different standard datasets: MNIST and CIFAR-10. The algorithm results are compared to the basic Adam algorithm and the SGD algorithm, in addition to other five recent SGD adaptive algorithms. In most comparisons, the HN Adam algorithm outperforms the compared algorithms in terms of accuracy and convergence speed. AdaBelief is the most competitive of the compared algorithms. In terms of testing accuracy and convergence speed (represented by the consumed training time), the HN-Adam algorithm outperforms the AdaBelief algorithm by an improvement of 1.0% and 0.29% for the MNIST dataset, and 0.93% and 1.68% for the CIFAR-10 dataset, respectively.

Nowadays, liver tumor segmentation has been widely applied for vital medical objectives such as illness diagnosis, treatment, and evaluation of liver function. In this work, we aim to improve the performance of hybrid convolution neural network (CNN) models for liver tumor segmentation. A new automatic design method is introduced to build a hybrid CNN model with an optimal architecture using a combination of the original U_Net and the Res-UNet. In the proposed design method, the Res-UNet model is divided into different blocks, and then many different candidate architectures for the hybrid CNN model can be created using gradually these blocks to replace their corresponding blocks of the original U-Net. The replacement is executed if the candidate architecture using a residual block gives better performance than the corresponding one that uses the original U-Net block. We use a genetic algorithm (GA) to get the optimal architecture among the candidate architectures that gives the best performance for the hybrid CNN model. During the optimization process, in addition to the optimal architecture, other crucial configurations such as the learning algorithm, learning rate, and batch size are also determined. We refer to the hybrid CNN model that has the best architecture as GA_Res_UNet. Comparison has been made with other CNN models that are commonly used in image segmentation of liver tumors. We use the 3D-IRCADb01 liver tumor dataset that contains 2800 magnetic resonance (MR) images of 20 patients. Each image has a liver tumor with its corresponding mask. All MR images are resized to 256 × 256 in DICOM format. The findings indicate that GA_Res_UNet outperformed the other compared models over the considered dataset. It achieves 98.5% of dice coefficient and 99.8% of accuracy. Moreover, two other liver tumor segmentation datasets (LiTS17 and CHAOS) are utilized in another comparison to show the superior performance of our proposed model. The reported results show that the proposed model outperformed the other compared state-of-the-art models.

The scarcity of water resources nowadays lays stress on researchers to develop strategies aiming at making the best benefit of the currently available resources. One of these strategies is ensuring that reliable and near-optimum designs of water distribution systems (WDSs) are achieved. Designing WDSs is a discrete combinatorial NP-hard optimization problem, and its complexity increases when more objectives are added. Among the many existing evolutionary algorithms, a new hybrid fast-convergent multi-objective particle swarm optimization (MOPSO) algorithm is developed to increase the convergence and diversity rates of the resulted non-dominated solutions in terms of network capital cost and reliability using a minimized computational budget. Several strategies are introduced to the developed algorithm, which are self-adaptive PSO parameters, regeneration-on-collision, adaptive population size, and using hypervolume quality for selecting repository members. A local search method is also coupled to both the original MOPSO algorithm and the newly developed one. Both algorithms are applied to medium and large benchmark problems. The results of the new algorithm coupled with the local search are superior to that of the original algorithm in terms of different performance metrics in the medium-sized network. In contrast, the new algorithm without the local search performed better in the large network.

Design of water distribution systems is commonly performed using design guidelines which often rely on designers' expertise. Determination of the required pumping power and the size and location of storage tanks, according to the network layout and the available diameters, is not a trivial task. The difficulty of optimizing such systems arises from applying non-uniform demand patterns along the one-day operation accompanied by the different constraints to satisfy the large fluctuations in pressure heads and flow velocities, the dilemma between choosing to reduce the cost of constructing and operating the network or to increase its resilience and the complex relationship between the pumping power and the size, bottom elevation and location of storage tanks. In this study, a general multi-objective optimization framework is developed to get the optimum diameters, pumping power and storage tank locations and bottom elevations to minimize the cost while maximizing the resilience using a combination of the Particle Swarm Optimization technique and EPANET software. The objective function of the optimization is obtained from combining the cost and resilience using a transformation function. The proposed framework is applied to Safi town network in Yemen and the results prove the success of performing the developed framework.

Optimizing the design of water distribution systems often faces difficulties due to continuous variations in water demands, pressure requirements, and disinfectant concentrations. The complexity of this optimization even increases when trying to optimize both the hydraulic and the water quality design models. Most of the previous works in the literature did not investigate the linkage between both models, either by combining them into one general model or by selecting any representative solution to proceed from one model to another. This work introduces an integrated two-step framework to optimize both designs while investigating the reasonable network configuration selection from the hydraulic design view before proceeding to the water quality design. The framework is mainly based on a modified version of the multi-objective particle swarm optimization algorithm. The algorithm’s first step is optimizing the hydraulic design of the network by minimizing the system’s capital cost while maximizing the system’s reliability. The second step targets optimizing the water quality design by minimizing both the total consumed chlorine mass and the accumulated differences between actual and maximum chlorine concentrations for all the network junctions. The framework is applied to Safi Network in Yemen. Three scenarios of the water quality design are proposed based on the selected decision variables. The results show a superior performance of the first scenario, based on optimized 24-h multipliers of a chlorine pattern for a flow-paced booster station, compared to the other scenarios in terms of the diversity of final solutions.