Type 2 Diabetes Risk Prediction Using Deep Convolutional Neural Network Based-Bayesian Optimization

Diabetes mellitus is a long-term condition characterized by hyperglycemia. It could lead to plenty of difficulties. According to rising morbidity in recent years, the world’s diabetic patients will exceed 642 million by 2040, implying that one out of every ten persons will be diabetic. There is no doubt that this startling figure requires immediate attention from industry and academia to promote innovation and growth in diabetes risk prediction to save individuals’ lives. Due to its rapid development, deep learning (DL) was used to predict numerous diseases. However, DL methods still suffer from their limited prediction performance due to the hyperparameters selection and parameters optimization. Therefore, the selection of hyper-parameters is critical in improving classification performance. This study presents Convolutional Neural Network (CNN) that has achieved remarkable results in many medical domains where the Bayesian optimization algorithm (BOA) has been employed for hyperparameters selection and parameters optimization. Two issues have been investigated and solved during the experiment to enhance the results. The first is the dataset class imbalance, which is solved using Synthetic Minority Oversampling Technique (SMOTE) technique. The second issue is the model's poor performance, which has been solved using the Bayesian optimization algorithm. The findings indicate that the Bayesian based-CNN model superbases all the state-of-the-art models in the literature with an accuracy of 89.36%, F1-score of 0.88.6, and Matthews Correlation Coefficient (MCC) of 0.88.6.     

GIS and Multi-Criteria Evaluation: Robust Tools for Integrated Water Resources Management

Abstract

Integrated water resources planning and management are considered very complex issues. These issues are usually addressed through the multi-sectoral, interdisciplinary and hierarchal decomposition approaches. In general, integrated resource management indicates the consideration of water, social, socio-economic, economic and environmental issues. The current study aims at merging the GIS and Multi-Criteria Evaluation (MCE) techniques for the integrated water resource management of a cropped area. An area of about 120,000 Hectares located in the Northern Nile River Delta area with a coastal zone on the Mediterranean was selected and GIS was applied to represent the area's different environment, social, economic, and water factors. Different randomization cropping pattern distribution scenarios were proposed. Through the merging of GIS and MCE approaches, three scenarios were run and evaluated at three different levels: farm, canal catchment area and whole area. This merging resulted in a very powerful tool for the evaluation of different plans. The merging of GIS and MCE really facilitated the decision making process for these types of integrated water management problems.     

Comparison of two quantitative proton density mapping methods in multiple sclerosis

Objective

Proton density (PD) mapping requires correction for the receive profile (RP), which is frequently performed via bias-field correction. An alternative RP-mapping method utilizes a comparison of uncorrected PD-maps and a value ρ(T1) directly derived from T1-maps via the Fatouros equation. This may be problematic in multiple sclerosis (MS), if respective parameters are only valid for healthy brain tissue. We aimed to investigate whether the alternative method yields correct PD values in MS patients.

Materials/methods

PD mapping was performed on 27 patients with relapsing-remitting MS and 27 healthy controls, utilizing both methods, yielding reference PD values (PD_ref, bias-field method) and PD_alt (alternative method).

Results

PD_alt-values closely matched PD_ref, both for patients and controls. In contrast, ρ(T1) differed by up to 3 % from PD_ref, and the voxel-wise correlation between PD_ref and ρ(T1) was reduced in a patient subgroup with a higher degree of disability. Still, discrepancies between ρ(T1) and PD_ref were almost identical across different tissue types, thus translating into a scaling factor, which cancelled out during normalization to 100 % in CSF, yielding a good agreement between PD_alt and PD_ref.

Conclusion

RP correction utilizing the auxiliary parameter ρ(T1) derived via the Fatouros equation provides accurate PD results in MS patients, in spite of discrepancies between ρ(T1) and actual PD values.

     

CFD simulation of the IAEA 10 MW generic MTR reactor under loss of flow transient

Three-dimensional simulation of the IAEA 10 MW generic reactor under loss of flow transient is introduced using the CFD code, Fluent. The IAEA reactor calculation is a safety-related benchmark problem for an idealized material testing reactor (MTR) pool type specified in order to compare calculational methods used in various research centers. The flow transients considered include fast loss of flow accidents (FLOFA) and slow loss of flow accidents (SLOFA) modeled with exponential flow decay and time constants of 1 and 25 s, respectively. The transients were initiated from a power of 12 MW with a flow trip point at 85% nominal flow and a 200 ms time delay. The simulation shows comparable results as those published by other research groups. However, interesting 3D patterns are shown that are usually lost based on the one-dimensional simulations that other research groups have introduced. In addition, information about the maximum clad surface temperature, the maximum fuel element temperature as well as the location of hot spots in fuel channel is also reported.     

Hydrogen refueling station compression and storage optimization with tube-trailer deliveries

Hydrogen refueling stations require high capital investment, with compression and storage comprising more than half of the installed cost of refueling equipment. Refueling station configurations and operation strategies can reduce capital investment while improving equipment utilization. Argonne National Laboratory developed a refueling model to evaluate the impact of various refueling compression and storage configurations and tube trailer operating strategies on the cost of hydrogen refueling. The modeling results revealed that a number of strategies can be employed to reduce fueling costs. Proper sizing of the high-pressure buffer storage reduces the compression requirement considerably, thus reducing refueling costs. Employing a tube trailer to initially fill the vehicle's tank also reduces the compression and storage requirements, further reducing refueling costs. Reducing the cut-off pressure of the tube trailer for initial vehicle fills can also significantly reduce the refueling costs. Finally, increasing the trailer's return pressure can cut refueling costs, especially for delivery distances less than 100 km, and in early markets, when refueling stations will be grossly underutilized.     

Location-Aware Personalized Traveler Recommender System (LAPTA) Using Collaborative Filtering KNN

Many tourists who travel to explore different cultures and cities worldwide aim to find the best tourist sites, accommodation, and food according to their interests. This objective makes it harder for tourists to decide and plan where to go and what to do. Aside from hiring a local guide, an option which is beyond most travelers’ budgets, the majority of sojourners nowadays use mobile devices to search for or recommend interesting sites on the basis of user reviews. Therefore, this work utilizes the prevalent recommender systems and mobile app technologies to overcome this issue. Accordingly, this study proposes location-aware personalized traveler assistance (LAPTA), a system which integrates user preferences and the global positioning system (GPS) to generate personalized and location-aware recommendations. That integration will enable the enhanced recommendation of the developed scheme relative to those from the traditional recommender systems used in customer ratings. Specifically, LAPTA separates the data obtained from Google locations into name and category tags. After the data separation, the system fetches the keywords from the user’s input according to the user’s past research behavior. The proposed system uses the K-Nearest algorithm to match the name and category tags with the user’s input to generate personalized suggestions. The system also provides suggestions on the basis of nearby popular attractions using the Google point of interest feature to enhance system usability. The experimental results showed that LAPTA could provide more reliable and accurate recommendations compared to the reviewed recommendation applications.     

Optimum VAR sizing and allocation using particle swarm optimization

This paper proposes a solution technique for finding the optimum location and sizing of the shunt compensation devices in transmission systems. The objective of the formulation is to improve the voltage stability of the system while maintaining acceptable voltage profile. The problem can be formulated as an integer nonlinear optimization problem. The newly developed evolutionary technique particle swarm optimization (PSO) is used to solve this problem. Case studies with the Ward–Hale 6-bus, IEEE 14-bus and 30-bus systems are presented to illustrate the applicability of the algorithm. A comparison with the genetic algorithm (GA) is performed to show the quality of the solutions obtained by PSO.     

On‐Chip Generation of Vortical Flows for Microfluidic Centrifugation

Microcentrifugation constitutes an important part of the microfluidic toolkit in a similar way that centrifugation is crucial to many macroscopic procedures, given that micromixing, sample preconcentration, particle separation, component fractionation, and cell agglomeration are essential operations in small scale processes. Yet, the dominance of capillary and viscous effects, which typically tend to retard flow, over inertial and gravitational forces, which are often useful for actuating flows and hence centrifugation, at microscopic scales makes it difficult to generate rotational flows at these dimensions, let alone with sufficient vorticity to support efficient mixing, separation, concentration, or aggregation. Herein, the various technologies—both passive and active—that have been developed to date for vortex generation in microfluidic devices are reviewed. Various advantages or limitations associated with each are outlined, in addition to highlighting the challenges that need to be overcome for their incorporation into integrated microfluidic devices.     

Ultrafast Acoustofluidic Exfoliation of Stratified Crystals

While the remarkable properties of 2D crystalline materials offer tremendous opportunities for their use in optics, electronics, energy systems, biotechnology, and catalysis, their practical implementation largely depends critically on the ability to exfoliate them from a 3D stratified bulk state. This goal nevertheless remains elusive, particularly in terms of a rapid processing method that facilitates high yield and dimension control. An ultrafast multiscale exfoliation method is reported which exploits the piezoelectricity of stratified materials that are noncentrosymmetric in nature to trigger electrically‐induced mechanical failure across weak grain boundaries associated with their crystal domain planes. In particular, it is demonstrated that microfluidic nebulization using high frequency acoustic waves exposes bulk 3D piezoelectric crystals such as molybdenum disulphide (MoS₂) and tungsten disulphide (WS₂) to a combination of extraordinarily large mechanical acceleration (≈10⁸ m s^?2) and electric field (≈10⁷ V m^?1). This results in the layered bulk material being rapidly cleaved into pristine quasi‐2D‐nanosheets that predominantly comprise single layers, thus constituting a rapid and high throughput chip‐scale method that opens new possibilities for scalable production and spray coating deposition.