Hybrid Computational Model Research Articles

Hybrid Computational Modeling for Web Application Security Assessment

Transformation from conventional business management systems to smart digital systems is a recurrent trend in the current era. This has led to digital revolution, and in this context, the hardwired technologies in the software industry play a significant role However, from the beginning, software security remains a serious issue for all levels of stakeholders. Software vulnerabilities lead to intrusions that cause data breaches and result in disclosure of sensitive data, compromising the organizations’ reputation that translates into, financial losses as well. Most of the data breaches are financially motivated, especially in the healthcare sector. The cyber invaders continuously penetrate the E-Health data because of the high cost of the data on the dark web. Therefore, security assessment of healthcare web-based applications demands immediate intervention mechanisms to weed out the threats of cyber-attacks. The aim of this work is to provide efficient and effective healthcare web application security assessment. The study has worked with the hybrid computational model of Multi-Criteria Decision Making (MCDM) based on Analytical Hierarchy Process (AHP) and Technique for Order of Preference by Similarity to Ideal-Solutions (TOPSIS) under the Hesitant Fuzzy (HF) environment. Hesitant fuzzy sets provide effective solutions to address decision making problems where experts counter hesitation to make a decision. The proposed research endeavor will support designers and developers in identifying, selecting and prioritizing the best security attributes for web applications’ development. The empirical analysis concludes that Robustness got highest priority amongst the assessed security attributes set followed by Encryption, Authentication, Limit Access, Revoke Access, Data Validation, and Maintain Audit Trail. The results of this research endeavor depict that this proposed computational procedure would be the most conversant mechanism for determining the web application security. The study also establishes guidelines which the developers can refer for the identification and prioritization of security attributes to build more secure and trustworthy web-based applications.

Multicellular mechanochemical hybrid cellular Potts model of tissue formation during epithelial‐mesenchymal transition

AbstractEpithelial‐mesenchymal transition (EMT) is the transdifferentiation of epithelial cells to a mesenchymal phenotype, in which cells lose epithelial‐like cell–cell adhesions and gain mesenchymal‐like enhanced contractility and mobility. EMT is crucial for tissue regeneration and is also implicated in pathological conditions, such as cancer metastasis. Prior work has shown that transforming growth factor‐1 (TGF‐1) is a potent inducer of this biological process. In this study, we develop a computational model coupling mechanical and biochemical signaling in a multicellular tissue undergoing EMT. Specifically, we utilize a recently developed formulation that integrates a multicellular cellular Potts model (CPM), a lattice‐based stochastic model governing cell movement; a first moment of area model, governing cellular traction and junctional forces; a finite element model, which defines extracellular matrix (ECM) substrate strains; an intracellular signaling TGF‐1‐mediated EMT model that governs cellular phenotype; and an extracellular signaling component governing ECM and TGF‐1 signaling. In this study, we modeled the spatial cellular patterns that occur in tissue and the ECM during EMT. Our model predicts that EMT often initially occurs at a tissue boundary due to mechanochemical coupling, which results in transdifferentiation to progress inwards toward the center. Variation in model parameters demonstrated conditions enhancing and suppressing EMT, especially to drive EMT in the absence of TGF‐1 and inhibit EMT in the presence of TGF‐1. Specifically, enhancing the mechanochemical feedback typically promoted EMT, whereas greater assembled ECM degradation suppressed EMT. Simulated scratch test experiments illustrate that ECM composition can impact closure directly through EMT signaling. In conclusion, we integrated mechanical, biochemical, and extracellular signaling networks in a novel hybrid computational model to reproduce tissue formation dynamics of EMT.

Hybrid Computational Model to Assist in the Location of Victims Buried in the Tragedy of Brumadinho

This paper presents a hybrid computational model based on regression techniques, machine learning and physicomathematical algorithms developed for assistance in locating victims in the Brumadinho tragedy in 2019. The physicomathematical model, which provided results to help search teams, is based on integral and vector calculus, and fluid mechanics concepts. In addition, from data provided by the physicomathematical algorithm, two hybrid model were developed. One of them uses regression statistical and the other one uses support vector regression which is a type of machine learning. With good prospects of the advances in research, it is expected in future work, a more accurate model that can be used in other possible situations of dam-break. Moreover the model can be applied to situations involving computational fluid dynamics in general

An efficient multi-level high-order algorithm for simulation of a class of Allen–Cahn stochastic systems

Computationally quantifying uncertainties in the mathematical modeling of physical processes is crucial for understanding the errors induced by both numerical approximations of the model and lack of precise input data assumed in the continuous model. Uncertain input parameters in the continuous model are typically treated as random variables, leading to the need to consider solutions of both the continuous and discrete models as stochastic processes. Computing statistical moments of the stochastic processes is an extremely important part of the uncertainty quantification problem. In this work, we consider a class of physical processes that are modeled by the Allen–Cahn (A–C) partial differential equation (PDE) evolutionary system, with uncertainties in the initial state of the evolution and the A–C PDE. We develop a hybrid computational model for the stochastic A–C system to efficiently compute statistical moments of the numerical counterparts of the A–C stochastic processes. The hybrid framework comprises finite element method in-space approximations, high-order digital nets based sampling in high dimensional probability space, and an interplay of discretization parameters in the spatial and stochastic approximations. We demonstrate marked efficiency of the hybrid framework, compared to the standard methods, using two- and three-dimensional in space and high stochastic dimensional A–C example systems.

Modeling Nonlinear Dendritic Processing of Facilitation in a Dragonfly Target-Tracking Neuron.

Dragonflies are highly skilled and successful aerial predators that are even capable of selectively attending to one target within a swarm. Detection and tracking of prey is likely to be driven by small target motion detector (STMD) neurons identified from several insect groups. Prior work has shown that dragonfly STMD responses are facilitated by targets moving on a continuous path, enhancing the response gain at the present and predicted future location of targets. In this study, we combined detailed morphological data with computational modeling to test whether a combination of dendritic morphology and nonlinear properties of NMDA receptors could explain these observations. We developed a hybrid computational model of neurons within the dragonfly optic lobe, which integrates numerical and morphological components. The model was able to generate potent facilitation for targets moving on continuous trajectories, including a localized spotlight of maximal sensitivity close to the last seen target location, as also measured during in vivo recordings. The model did not, however, include a mechanism capable of producing a traveling or spreading wave of facilitation. Our data support a strong role for the high dendritic density seen in the dragonfly neuron in enhancing non-linear facilitation. An alternative model based on the morphology of an unrelated type of motion processing neuron from a dipteran fly required more than three times higher synaptic gain in order to elicit similar levels of facilitation, despite having only 20% fewer synapses. Our data support a potential role for NMDA receptors in target tracking and also demonstrate the feasibility of combining biologically plausible dendritic computations with more abstract computational models for basic processing as used in earlier studies.

Experimental and computational analysis of bending fatigue failure in chopped carbon fiber chip reinforced composites

With a better balance among good mechanical performance, high freedom of design, and low material and manufacturing cost, chopped carbon fiber chip reinforced sheet molding compound (SMC) composites show great potential in different engineering applications. In this paper, bending fatigue behaviors of SMC composites considering the heterogeneous fiber orientation distributions have been thoroughly investigated utilizing both experimental and computational methods. First, four-point bending fatigue tests are performed with designed SMC composites, and the local modulus is adopted as a metric to represent the local fiber orientation of two opposing sides. Interestingly, SMC composites with and without large discrepancy in local modulus of opposing sides show different fatigue behaviors. Interrupted tests are conducted to explore the bending fatigue failure mechanism, and the damage processes of valid specimens are also closely examined. We find that the fatigue failure of SMC composites under four-point bending is governed by crack propagation instead of crack initiation. Because of this, the heterogeneous local fiber orientations of both sides of the specimen influence fatigue life. The microstructure of the lower side shows a direct influence while that of the upper side also exhibiting influence which becomes more prominent for high cycle fatigue cases. Furthermore, a hybrid micro–macro computational model is proposed to efficiently study the cyclic bending behavior of SMC composites. The region of interest is reconstructed with a modified random sequential absorption algorithm to conserve all the microstructural details including the heterogeneous fiber orientation, while the rest of the regions are modeled as homogenized macro-scale continua. Combined with a framework to capture the progressive fatigue damage under cyclic bending, the bending fatigue behaviors of SMC composites are accurately captured by the hybrid computational model comparing with our experimental analysis.

A Spatial Quantitative Systems Pharmacology Platform spQSP-IO for Simulations of Tumor-Immune Interactions and Effects of Checkpoint Inhibitor Immunotherapy.

Simple SummaryMathematical and computational models, such as quantitative systems pharmacology (QSP) models, are becoming a popular tool in drug discovery. The advancement of imaging techniques creates a unique opportunity to further expand these models and enhance their predictive power by incorporating characteristics of individual patients embedded in the image data obtained from tissue samples. The aim of this study is to develop a platform which combines the strength of QSP models and spatially resolved agent-based models (ABM), and create a model of cancer development in the context of anti-cancer immunity and immune checkpoint inhibition therapy. The model can be applied in virtual clinical trials and biomarker discovery to help refine trial design.Quantitative systems pharmacology (QSP) models have become increasingly common in fundamental mechanistic studies and drug discovery in both academic and industrial environments. With imaging techniques widely adopted and other spatial quantification of tumor such as spatial transcriptomics gaining traction, it is crucial that these data reflecting tumor spatial heterogeneity be utilized to inform the QSP models to enhance their predictive power. We developed a hybrid computational model platform, spQSP-IO, to extend QSP models of immuno-oncology with spatially resolved agent-based models (ABM), combining their powers to track whole patient-scale dynamics and recapitulate the emergent spatial heterogeneity in the tumor. Using a model of non-small-cell lung cancer developed based on this platform, we studied the role of the tumor microenvironment and cancer–immune cell interactions in tumor development and applied anti-PD-1 treatment to virtual patients and studied how the spatial distribution of cells changes during tumor growth in response to the immune checkpoint inhibition treatment. Using parameter sensitivity analysis and biomarker analysis, we are able to identify mechanisms and pretreatment measurements correlated with treatment efficacy. By incorporating spatial data that highlight both heterogeneity in tumors and variability among individual patients, spQSP-IO models can extend the QSP framework and further advance virtual clinical trials.

3D hybrid model of foundation‐soil‐foundation dynamic interaction

Abstract3D dynamic interaction of two adjacent elastic foundations embedded in a finite layered soil region rested in a homogeneous elastic isotropic half‐space with a transient dynamic source is studied. The hybrid computational model, the corresponding numerical scheme and the accompanied software are developed, verified and inserted in detail parametric study. The hybrid approach is based on (a) finite element method (FEM) describing the scattered wave field in a finite layered soil profile with two foundations; (b) boundary element method (BEM) considering waves radiating from a dynamic transient source in elastic semi‐infinite range. The aim is to propose an efficient hybrid methodology for evaluation of the dynamic response of a foundation‐soil‐foundation system, taking into account (a) the whole wave path from the dynamic source, through the layered soil region, till the underground structures; (b) the damage state of the geological material. The BEM model is based on the 3D elastodynamic fundamental solution in Fourier domain and it is applied in order to obtain frequency‐dependent stiffness matrix and load vector of the dynamically active semi‐infinite geological zone. Once the BEM model is formulated, it is inserted as a macro‐finite element in the FEM software package ABAQUS. The frequency‐dependent FEM model describing the wave field in finite layered soil profile with two elastic foundations is realized by ABAQUS. Solutions in time domain are obtained through application of the inverse fast Fourier transform. As a final result an efficient hybrid model comprising all in one: dynamic (seismic or other type) source‐homogeneous elastic half‐space‐finite heterogeneous by layers and foundations soil profile, is presented. Parametric study illustrating the sensitivity of the dynamic field to key factors such as the type and properties of the load, the soil layering, the material properties of the far‐ and near‐field soil regions, the damaged state of the geological material and the foundation‐soil‐foundation interaction is shown and discussed.

Reconstructing long-term wind speed data based on measure correlate predict method for micro-grid planning

One of the most significant uncertain parameters in microgrid planning is wind speed. Wind speed has a complex dynamic behavior and developing models with high accuracy for this parameter can help decrease microgrid costs and can help improve reliability. Previous methods for wind speed modeling have been mostly statistical methods (e.g., the auto-regressive integrated moving average (ARIMA), the multivariate distribution functions, Copula function). Afore-mentioned methods are suitable for modeling stationary data. However, since the wind speed data are not stationary, the measure correlate predict (MCP) method, capable of modeling the non-stationary data, is used in this paper to generate the wind speed scenarios. Based on radial basis function (RBF) artificial neural network, this work has proposed a novel hybrid computational model to improve MCP method’s performance. The results indicate that the Hybrid-MCP method is more accurate than the conventional statistical methods, which are used to generate the wind speed scenarios (24 h).

Expansion dynamics and radiation absorption in plumes induced by irradiation of a copper target by single and multiple nanosecond laser pulses in the doughnut beam mode

The two-dimensional kinetic simulations of the plume expansion process induced by irradiation of a copper target in argon background gas with nanosecond laser pulses in the doughnut beam mode are performed. The target is irradiated by single and double laser pulses, as well as by bursts consisting of up to 35 individual pulses. The simulations are based on a hybrid computational model that includes the direct simulation Monte Carlo method, where the ionization and absorption of laser radiation are considered. The numerical results are systematically compared with the results obtained for a Gaussian beam at the same pulse energy, peak fluence, and peak intensity. The structure of the plumes induced by a doughnut beam is found to drastically differ from the structure of the plumes induced by a Gaussian beam. The doughnut beam initially induces a toroidal shock wave. The self-interaction of this shock at the axis of symmetry results in the formation of multiple internal shock waves that strongly increase the plume temperature and density. The interaction between plumes induced by a double pulse in the doughnut beam mode occurs in two distinct regimes depending on the beam size and inter-pulse separation time. At a relatively small peak laser fluence, the effect of plasma shielding in the plumes induced by a burst of pulses in the doughnut beam mode is found to be smaller compared to the case of a Gaussian beam. At a relatively large fluence, the simulations reveal the opposite trend, but the amount of vaporized material is found to be always larger for a doughnut beam than for a Gaussian one. The obtained results suggest that the absorption of laser radiation in plasma plumes induced by nanosecond laser pulses can be tuned by changing the radial distribution of energy in the incident laser beam.

Nucleation Rate Prediction of Curcumin Particles in Microfluidic‐Assisted Nanoprecipitation

AbstractA hybrid computational model was developed to interpret the formation of curcumin particles via liquid antisolvent precipitation (LASP) in three microchannel reactors (MCRs) with different confluence angles. The computational fluid dynamics (CFD) model was numerically combined with a discretized population balance approach. The nucleation kinetics were computed using an expression obtained from the nucleation rate and supersaturation equation from previous experimental data. A quantitative comparison of the particle size distributions (PSDs) demonstrated that the particles with the smallest diameters are observed right after the intersection of the three streams. Besides, the large confluence angle of 135° in the MCR presented higher population density as a result of the smaller size of the precipitated curcumin particles. These findings illustrate the efficient applicability of the hybrid model.

GIS-based evaluation of landslide susceptibility using a novel hybrid computational intelligence model on different mapping units

Landslide susceptibility mapping is significant for landslide prevention. Many approaches have been used for landslide susceptibility prediction, however, their performances are unstable. This study constructed a hybrid model, namely box counting dimension-based kernel logistic regression model, which uses fractal dimension calculated by box counting method as input data based on grid cells mapping unit and terrain mapping unit. The performance of this model was evaluated in the application in Zhidan County, Shaanxi Province, China. Firstly, a total of 221 landslides were identified and mapped, and 11 landslide predisposing factors were considered. Secondly, the landslide susceptibility maps (LSMs) of the study area were obtained by constructing the model on two different mapping units. Finally, the results were evaluated with five statistical indexes, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and Accuracy. The statistical indexes of the model obtained on the terrain mapping unit were larger than those based on grid cells mapping unit. For training and validation datasets, the area under the receiver operating characteristic curve (AUC) of the model based on terrain mapping unit were 0.9374 and 0.9527, respectively, indicating that establishing this model on the terrain mapping unit was advantageous in the study area. The results show that the fractal dimension improves the prediction ability of the kernel logistic model. In addition, the terrain mapping unit is a more promising mapping unit in Loess areas.

Deep Learning Based Hybrid Computational Intelligence Models for Options Pricing

Options are commonly used by traders and investors for hedging their investments. They also allow the traders to execute leveraged trading opportunities. Meanwhile accurately pricing the intended option is crucial to perform such tasks. The most common technique used in options pricing is Black–Scholes (BS) formula. However, there are slight differences between the BS model output and the actual options price due to the ambiguity in defining the volatility. In this study, we developed hybrid deep learning based options pricing models to achieve better pricing compared to BS. The results indicate that the proposed models can generate more accurate prices for all option classes. Compared with BS using annualized 20 intraday returns as volatility, 94.5% improvement is achieved in option pricing in terms of mean squared error.

Neuro-Symbolic Integration of Hopfield Neural Network for Optimal Maximum Random kSatisfiability (Maxrksat) Representation

Boolean satisfiability logical representation is a programming paradigm that has its foundations in mathematical logic. It has been classified as an NP-complete problem that difficult practical combinatorial optimization and search problems can be easily converted into it. Random Maximum kSatisfiability (MAX-RkSAT) composed of the most consistent mapping in a Boolean formula that generates a maximum number of random satisfied clauses. Many optimization and search problems can be easily expressed by mapping the problem into a Hopfield neural network (HNN) to minimize the optimal configuration of the corresponding Lyapunov energy function. In this paper, a hybrid computational model hs been proposed that incorporates the Random Maximum kSatisfiability (MAX-RkSAT) into the Hopfield neural network (HNN) for optimal Random Maximum kSatisfiability representation (HNN-MAX-RkSAT). Hopfield neural network learning will be integrated with the random maximum satisfiability to enhance the correct neural state of the network model representation. The computer simulation using C++ has been used to demonstrate the ability of MAX-RkSAT to be embedded optimally in Hopfield neural network to serve as Neuro-symbolic integration. The performance of the proposed hybrid HNN-MAXRkSAT model has been explored and compared with the existing model. The proposed HNN-MAXRkSAT demonstrates good agreement with the existing models measured in terms of Global minimum Ratio (Gm), Hamming Distance (HD), Mean Absolute Error (MAE) and network computation Time CPU time). The proposed framework explored that MAX-RkSAT can be optimally represented in HNN and subsequently provides an additional platform for neural-symbolic integration, representing the various types of satisfiability logic.

GIS-based evaluation of landslide susceptibility using hybrid computational intelligence models

Landslides have caused huge economic and human losses in China. Mapping of landslide susceptibility is an important tool to prevent and control landslide disasters. The purpose of this study is to make use of a hybrid machine learning approach by combining the reduced-error pruning trees (Rept) with a series of ensemble techniques (Bagging, Dagging, and Real Adaboost) and compare the performance of each combination for landslide susceptibility modeling. The combination of Rept model and Real Adaboost (RRept)is a novel application in the field of landslide susceptibility. Firstly, a landslide inventory map was prepared with 298 determined historical landslides events in the study area, 209 landslides (70%) were randomly selected for the training dataset and the remaining 89 landslides (30%) were used for validation dataset. On this basis, 16 landslide influencing factors were included in the landslide susceptibility evaluation (slope angle, elevation, slope aspect, sediment transport index (STI), topographical wetness index (TWI), stream power index (SPI), profile curvature, plan curvature, distance to rivers, distance to roads, distance to faults, soil, normalized difference vegetation index (NDVI), landuse, lithology and rainfall). Secondly, the correlation attribute evaluation (CAE) method was used to select the most important factors for the proposed landslide susceptibility model. The results show that all the factors contribute to the occurrence of landslide. Slope angle, Landuse, Elevation, Distance to roads, Soil and Lithology have the greatest influence on the occurrence of landslide. The receiver operating characteristics (ROC), standard error (SE), 95% confidence interval and mean absolute error (MAE) were then used to validate and compare the performance of the model. The best model should have the largest AUC value, the smallest SE, the narrowest 95% CI and the smallest MAE. The results show that the three hybrid models perform better than the Rept model alone. For the training data set, the RRept model has highest AUC value (0.927), the smallest SE (0.121), the narrowest 95% confidence interval (0.898–0.95) and the lowest MAE (0.20). For the validation data set, the RRept model has the highest AUC value (0.745), the narrowest 95% confidence interval (0.674–0.807) and the lowest MAE (0.33). The RRept has the highest predictive power for landslide susceptibility evaluation. The results show that a hybrid method improves the prediction ability of the base Rept model. In addition, the RRept model is a promising comprehensive model that can be applied to landslide susceptibility mapping.

Effect of fracture roughness on the hydrodynamics of proppant transport in hydraulic fractures

The effect of fracture roughness is investigated on proppant transport in hydraulic fractures using Joint Roughness Coefficient and a three-dimensional multiphase modelling approach. The equations governing the proppant transport physics in the fracturing fluid is solved using the hybrid computational fluid dynamics model. The reported proppant transport models in the literature are limited to the assumption of a smooth fracture domain with no fluid leak-off or fluid flow from fracture to rock matrix interface. In this paper, a proppant transport model is proposed that accounts for the proppant distribution in rough fracture geometry with fluid leak-off effect to surrounding porous rock. The hydrodynamic and mechanical behaviour of proppant transport was found directly related to the fracture roughness and flow regime especially under the influence of low viscosity fracturing fluid typically used in shale gas reservoirs. For the proppant transport in smooth fractures, the fracture walls employ mechanical retardation effects and reduce the proppant horizontal velocity resulting in more significant proppant deposition. On the contrary, for the proppant transport in rough fractures, the inter-proppant and proppant wall interactions become dominant that adds turbulence to the flow. It results in mechanical interaction flow effects becoming dominant and consequently higher proppants suspended in the slurry and greater horizontal transport velocity. Furthermore, the mechanical interaction flow effects were found to be principally dependant on the proppant transport regime and become significant at higher proppant Reynolds number.

Hybrid Model of Bacterial Biofilm Growth.

Bacterial biofilms play a critical role in environmental processes, water treatment, human health, and food processing. They exhibit highly complex dynamics due to the interactions between the bacteria and the extracellular polymeric substance (EPS), water, and nutrients and minerals that make up the biofilm. We present a hybrid computational model in which the dynamics of discrete bacterial cells are simulated within a multiphase continuum, consisting of EPS and water as separate interacting phases, through which nutrients and minerals diffuse. Bacterial cells in our model consume water and nutrients in order to grow, divide, and produce EPS. Consequently, EPS flows outward from the bacterial colony, while water flows inward. The model predicts bacterial colony formation as a treelike structure. The distribution of bacterial growth and EPS production is found to be sensitive to the pore spacing between bacteria and the consumption of nutrients within the bacterial colony. Forces that are sometimes neglected in biofilm simulations, such as lubrication force between nearby bacterial cells and osmotic (swelling) pressure force resulting from gradients in EPS concentration, are observed to have an important effect on biofilm growth via their influence on bacteria pore spacing and associated water/nutrient percolation into the bacterial colony.

Hybrid modeling of alpha rhythm and the amplitude of low-frequency fluctuations abnormalities in the thalamocortical region and basal ganglia in Alzheimer's disease.

A hybrid computational model of thalamocortical circuitry and basal ganglia is proposed to investigate the relation between the fractional amplitude of low-frequency fluctuations (fALFF) in the resting-state functional magnetic resonance imaging (rs-fMRI) in the striatum and electroencephalogram (EEG) changes within the alpha frequency bands in thalamic region in the case of Alzheimer's disease (AD). For that purpose, an Izikhevich neuron model-based network of the basal ganglia region is constructed and connected with the thalamic region which is modeled as neural mass. By considering the neurodegenerative changes in AD, the network dynamics are analyzed. The relation between the neural activity of basal ganglia and AD is investigated by modeling the blood oxygenation level-dependent (BOLD) signal. Decrease in fALFF of slow-4 band in the simulated BOLD signal of the striatum is observed. As the thalamic region receives inhibitory connections from basal ganglia over globus pallidus internal segment (GPi), the parameter changes emulating AD degenerations in the striatum increased the inhibitory effect on the thalamic network, and as a result, slowing in alpha rhythms is observed. It is observed that the decrease in the synaptic strength between the neurons in the striatum has a dominant effect on the slowing in alpha rhythm and also causes a decrease in fALFF of slow-4 band in striatum. This demonstrates a close and causal relation between the decrease in fALFF in the striatum and the slowing in alpha rhythms in the thalamic region in AD.

Evolutionary dynamics of cancer: From epigenetic regulation to cell population dynamics—mathematical model framework, applications, and open problems

Predictive modeling of the evolutionary dynamics of cancer is a challenging issue in computational cancer biology. In this paper, we propose a general mathematical model framework for the evolutionary dynamics of cancer, including plasticity and heterogeneity in cancer cells. Cancer is a group of diseases involving abnormal cell growth, during which abnormal regulation of stem cell regeneration is essential for the dynamics of cancer development. In general, the dynamics of stem cell regeneration can be simplified as a G0 phase cell cycle model, which leads to a delay differentiation equation. When cell heterogeneity and plasticity are considered, we establish a differential-integral equation based on the random transition of epigenetic states of stem cells during cell division. The proposed model highlights cell heterogeneity and plasticity; connects the heterogeneity with cell-to-cell variance in cellular behaviors (for example, proliferation, apoptosis, and differentiation/senescence); and can be extended to include gene mutation-induced tumor development. Hybrid computational models are developed based on the mathematical model framework and are applied to the processes of inflammation-induced tumorigenesis and tumor relapse after chimeric antigen receptor (CAR)-T cell therapy. Finally, we propose several mathematical problems related to the proposed differential-integral equation. Solutions to these problems are crucial for understanding the evolutionary dynamics of cancer.

Hybrid Computational Intelligence Models for Improvement Gully Erosion Assessment

Gullying is a type of soil erosion that currently represents a major threat at the societal scale and will likely increase in the future. In Iran, soil erosion, and specifically gullying, is already causing significant distress to local economies by affecting agricultural productivity and infrastructure. Recognizing this threat has recently led the Iranian geomorphology community to focus on the problem across the whole country. This study is in line with other efforts where the optimal method to map gully-prone areas is sought by testing state-of-the-art machine learning tools. In this study, we compare the performance of three machine learning algorithms, namely Fisher’s linear discriminant analysis (FLDA), logistic model tree (LMT) and naïve Bayes tree (NBTree). We also introduce three novel ensemble models by combining the aforementioned base classifiers to the Random SubSpace (RS) meta-classifier namely RS-FLDA, RS-LMT and RS-NBTree. The area under the receiver operating characteristic (AUROC), true skill statistics (TSS) and kappa criteria are used for calibration (goodness-of-fit) and validation (prediction accuracy) datasets to compare the performance of the different algorithms. In addition to susceptibility mapping, we also study the association between gully erosion and a set of morphometric, hydrologic and thematic properties by adopting the evidential belief function (EBF). The results indicate that hydrology-related factors contribute the most to gully formation, which is also confirmed by the susceptibility patterns displayed by the RS-NBTree ensemble. The RS-NBTree is the model that outperforms the other five models, as indicated by the prediction accuracy (area under curve (AUC) = 0.898, Kappa = 0.748 and TSS = 0.697), and goodness-of-fit (AUC = 0.780, Kappa = 0.682 and TSS = 0.618). The analyses are performed with the same gully presence/absence balanced modeling design. Therefore, the differences in performance are dependent on the algorithm architecture. Overall, the EBF model can detect strong and reasonable dependencies towards gully-prone conditions. The RS-NBTree ensemble model performed significantly better than the others, suggesting greater flexibility towards unknown data, which may support the applications of these methods in transferable susceptibility models in areas that are potentially erodible but currently lack gully data.