Multilayer Perceptron Training Research Articles

Exploring Gut Microbiome in Predicting the Efficacy of Immunotherapy in Non-Small Cell Lung Cancer.

We performed various analyses on the taxonomic and functional features of the gut microbiome from NSCLC patients treated with immunotherapy to establish a model that may predict whether a patient will benefit from immunotherapy. We collected 65 published whole metagenome shotgun sequencing samples along with 14 samples from our previous study. We systematically studied the taxonomical characteristics of the dataset and used both the random forest (RF) and the multilayer perceptron (MLP) neural network models to predict patients with progression-free survival (PFS) above 6 months versus those below 3 months. Our results showed that the RF classifier achieved the highest F-score (85.2%) and the area under the receiver operating characteristic curve (AUC) (95%) using the protein families (Pfam) profile, and the MLP neural network classifier achieved a 99.9% F-score and 100% AUC using the same Pfam profile. When applying the model trained in the Pfam profile directly to predict the treatment response, we found that both trained RF and MLP classifiers significantly outperformed the stochastic predictor in F-score. Our results suggested that such a predictive model based on functional (e.g., Pfam) rather than taxonomic profile might be clinically useful to predict whether an NSCLC patient will benefit from immunotherapy, as both the F-score and AUC of functional profile outperform that of taxonomic profile. In addition, our model suggested that interactive biological processes such as methanogenesis, one-carbon, and amino acid metabolism might be important in regulating the immunotherapy response that warrants further investigation.

Application of deep learning and spectral deconvolution for estimating mineral abundances of zeolite, Mg-sulfate and montmorillonite mixtures and its implications for Mars

A technique for estimating clinoptilolite, montmorillonite, and epsomite mineral abundances from a reflectance spectrum of mineral mixtures using spectral deconvolution and a deep neural network is proposed. Sixty-six ternary mineral mixtures were physically prepared with <150 μ m grain size with different weight percentages of minerals. A combination of normal and skewed Gaussian curves was fitted to the absorption bands at 1.4 μ m , 1.9 μ m , and 2.2 μ m of the acquired reflectance spectra of these mineral mixtures. Six Gaussian curve parameters with maximum absorption band depth ∼1.9 μ m , and wavelength at the maximum band depth, were used (along with mineral abundances) to train multilayer perceptron deep neural network (MLP-DNN) models. Forty-eight models with different DNN architectures and different hyperparameters were trained and the results were validated to find the best models. Winning models were tested using twenty-five samples including fourteen library spectra from RELAB and USGS spectral databases, a spectrum from a different sample, five different amounts of noise-added spectra simulating CRISM (Compact Reconnaissance Imaging Spectrometer for Mars) orbital spectral data, and five mixed spectra derived from linear mixtures of laboratory minerals. The best model was able to predict mineral mixture composition with higher accuracies; three of five montmorillonite and four of the five epsomite library spectra were identified with more than 90% accuracy. However, clinoptilolite samples show less than 50% accuracy and always predicted as mixtures of clinoptilolite and epsomite. This shows the difficulties of distinguishing non-analcime zeolites (e.g., clinoptilolite) from Mg-polyhydrated sulfate minerals, as discussed by authors who mapped hydrous minerals on Mars using hyperspectral image data. Also, the presence of at least ∼10% of montmorillonite in a clinoptilolite-montmorillonite mixture can entirely mask the presence of clinoptilolite in SWIR spectral data. Random artifacts introduced by noise sometime lead to predictions of completely different and incorrect mineral abundances. The study also discusses the possible reasons for the incorrect prediction of mineral abundances and how to overcome these difficulties. Overall, the study shows the advantage of spectral deconvolution with deep neural network for calculating mineral abundance from mixed mineral spectra. • Deep neural network estimating mineral abundances from spectroscopy is proposed. • Presence of montmorillonite mask the spectral identification of clinoptilolite. • Montmorillonite will also mask the presence of epsomite in SWIR spectral data.

Evaluating Three Supervised Machine Learning Algorithms (LM, BR, and SCG) for Daily Pan Evaporation Estimation in a Semi-Arid Region

Evaporation is one of the main components of the hydrological cycle, and its estimation is crucial and important for water resources management issues. Access to a reliable estimator tool for evaporation simulation is important in arid and semi-arid areas such as Iran, which lose more than 70% of their received precipitation by evaporation. Current research employs the Bayesian Regularization (BR) and Scaled Conjugate Gradient (SCG) algorithms for training the Multilayer Perceptron (MLP) model (as MLP-BR and MLP-SCG) and comparing their performance with the Levenberg–Marquardt (LM) algorithm (as MLP-LM). For this purpose, 16 meteorological variables were used on a daily scale; including temperature (5 variables), air pressure (4 variables), and relative humidity (6 variables) as input data sets, and pan evaporation as the target variable of the MLP model. The surveys were conducted during the period of 2006–2021 in Fars Province in Iran, which is a semi-arid region and has many natural lakes. Various combinations of input-target pairs were tested by several learning algorithms, resulting in seven input scenarios: (1) temperature-based (T), (2) pressure-based (F), (3) humidity-based (RH), (4) temperature–pressure-based (T-F), (5) temperature–humidity-based (T-RH), (6) pressure–humidity-based (F-RH) and (7) temperature–pressure–humidity-based (T-F-RH). The results indicated the relative superiority of the three-component scenario of T-F-RH, and a considerable weakness in the single-component scenario of RH compared with others. The best performance with a root mean square error (RMSE) equal to 1.629 and 1.742 mm per day and a Wilmott Index (WI) equal to 0.957 and 0.949 (respectively for validation and test periods) belonged to the MLP-BR model. Additionally, the amount of R2 (greater than 84%), Nash-Sutcliff efficiency (greater than 0.8) and normalized RMSE (less than 0.1) all indicate the reliability of the estimates provided for the daily pan evaporation. In the comparison between the studied training algorithms, two algorithms, BR and SCG, in most cases, showed better performance than the powerful and common LM algorithm. The obtained results suggest that future researchers in this field consider BR and SCG training algorithms for the supervised training of MLP for the numerical estimation of pan evaporation by the MLP model.

A NEW METHOD FOR DEHAZING OF UAV REMOTE SENSING IMAGES BASED ON IMPROVED DARK CHANNEL PRIOR

Abstract. Aiming to solve the problem of loss of important image information, such as blurred details and low contrast, caused by fog, haze and other meteorological influences in the slope monitoring process of UAV remote sensing images, a new method of improving dark channel image dehazing based on channel-weighted analysis and compensation function is proposed by a training of multilayer perceptron (MLP) in this study. First, based on the dark channel prior principle, the original hazy UAV image is mapped to obtain the estimated values of atmospheric light and rough transmittance. Next, by counting the RGB three-channel values of the pixels in the high-brightness regions, and analyzing the scattering of the RGB three-channel values in the haze, a color recovery module of atmospheric light is constructed, and the estimated value of atmospheric light is optimized. Then, according to the global transmittance, the compensation boundary value is determined and a functional relationship between different brightness regions and the increment of transmittance is established as a compensation function to optimize the rough transmittance. Finally, perform secondary optimization on the rough transmittance with the multi-layer perceptron (MLP) to obtain a smoother transmittance value. The experimental results show that the image processed by the proposed method has good contrast. The color saturation and authenticity are effectively maintained. And the detailed information of the mountain recorded by the image is better restored, which can provide a real data basis for slope monitoring.

Classification of Marine Mammals Using the Trained Multilayer Perceptron Neural Network with the Whale Algorithm Developed with the Fuzzy System.

The existence of various sounds from different natural and unnatural sources in the deep sea has caused the classification and identification of marine mammals intending to identify different endangered species to become one of the topics of interest for researchers and activist fields. In this paper, first, an experimental data set was created using a designed scenario. The whale optimization algorithm (WOA) is then used to train the multilayer perceptron neural network (MLP-NN). However, due to the large size of the data, the algorithm has not determined a clear boundary between the exploration and extraction phases. Next, to support this shortcoming, the fuzzy inference is used as a new approach to developing and upgrading WOA called FWOA. Fuzzy inference by setting FWOA control parameters can well define the boundary between the two phases of exploration and extraction. To measure the performance of the designed categorizer, in addition to using it to categorize benchmark datasets, five benchmarking algorithms CVOA, WOA, ChOA, BWO, and PGO were also used for MLPNN training. The measured criteria are concurrency speed, ability to avoid local optimization, and the classification rate. The simulation results on the obtained data set showed that, respectively, the classification rate in MLPFWOA, MLP-CVOA, MLP-WOA, MLP-ChOA, MLP-BWO, and MLP-PGO classifiers is equal to 94.98, 92.80, 91.34, 90.24, 89.04, and 88.10. As a result, MLP-FWOA performed better than other algorithms.

Ensembles of Biologically Inspired Optimization Algorithms for Training Multilayer Perceptron Neural Networks

Artificial neural networks have proven to be effective in a wide range of fields, providing solutions to various problems. Training artificial neural networks using evolutionary algorithms is known as neuroevolution. The idea of finding not only the optimal weights and biases of a neural network but also its architecture has drawn the attention of many researchers. In this paper, we use different biologically inspired optimization algorithms to train multilayer perceptron neural networks for generating regression models. Specifically, our contribution involves analyzing and finding a strategy for combining several algorithms into a hybrid ensemble optimizer, which we apply for the optimization of a fully connected neural network. The goal is to obtain good regression models for studying and making predictions for the process of free radical polymerization of methyl methacrylate (MMA). In the first step, we use a search procedure to find the best parameter values for seven biologically inspired optimization algorithms. In the second step, we use a subset of the best-performing algorithms and improve the search capability by combining the chosen algorithms into an ensemble of optimizers. We propose three ensemble strategies that do not involve changes in the logic of optimization algorithms: hybrid cascade, hybrid single elite solution, and hybrid multiple elite solutions. The proposed strategies inherit the advantages of each individual optimizer and have faster convergence at a computational effort very similar to an individual optimizer. Our experimental results show that the hybrid multiple elite strategy ultimately produces neural networks which constitute the most dependable regression models for the aforementioned process.

Lévy flight and chaos theory based gravitational search algorithm for multilayer perceptron training

Lévy flight and chaos theory based gravitational search algorithm for multilayer perceptron training

Degrees of freedom for off-the-grid sparse estimation

The degrees of freedom quantify the number of parameters of an estimator and is central to various risk minimization procedures. Its computation and properties are well understood when dealing with finite dimensional linear models, possibly regularized using sparsity through the celebrated Lasso estimator. However, for some applications (such as super-resolution in imaging or training multi-layer perceptrons with a single hidden layer) it makes sense to rather consider “continuous” methods. For these applications, these methods avoid the discretization of the parameter space and lead to more efficient numerical solvers. Training these continuous models with a sparsity inducing prior can be achieved by solving a convex optimization problem over the infinite dimensional space of measures, which is often called the Beurling Lasso (Blasso), and is the continuous counterpart of the Lasso. Previous works (Ann. Statist. 35 (2007) 2173–2192; Statist. Sinica 23 (2013) 809–828) show that the size of the smallest solution support is an unbiased estimator of the degrees of freedom of the Lasso. Our main contribution is a proof of a continuous counterpart to this result for the Blasso. In contrast to the Lasso, our new formula shows that the empirical degrees of freedom of the Blasso is not proportional to the size of the support. Our second contribution is a detailed study of the case of sampling Fourier coefficients in 1D, which corresponds to a super-resolution problem. We show that our formula for the unbiased estimation of the degrees of freedom is valid outside of a set of measure zero of observations, which in turn justifies its use to compute an unbiased estimator of the prediction risk using the Stein Unbiased Risk Estimator (SURE). We also report numerical results for both the case of Fourier sampling and the learning of a multilayers perceptron with a single hidden layer.

MLP-PSO Hybrid Algorithm for Heart Disease Prediction.

Background: Machine Learning (ML) is becoming increasingly popular in healthcare, particularly for improving the timing and accuracy of diagnosis. ML can provide disease prediction by analyzing vast amounts of healthcare data, thereby, empowering patients and healthcare providers with information to make informed decisions about disease prevention. Due to the rising cost of treatment, one of the most important topics in clinical data analysis is the prediction and prevention of cardiovascular disease. It is difficult to manually calculate the chances of developing heart disease due to a myriad of contributing factors. Objective: The aim of this paper is to develop and compare various intelligent systems built with ML algorithms for predicting whether a person is likely to develop heart disease using the publicly available Cleveland Heart Disease dataset. This paper describes an alternative multilayer perceptron (MLP) training technique that utilizes a particle swarm optimization (PSO) algorithm for heart disease detection. Methods: The proposed MLP-PSO hybrid algorithm and ten different ML algorithms are used in this study to predict heart disease. Various classification metrics are used to evaluate the performance of the algorithms. Results: The proposed MLP-PSO outperforms all other algorithms, obtaining an accuracy of 84.61%. Conclusions: According to our findings, the current MLP-PSO classifier enables practitioners to diagnose heart disease earlier, more accurately, and more effectively.

Modelling rate of penetration in drilling operations using RBF, MLP, LSSVM, and DT models

One of the most important problems that the drilling industry faces is drilling cost. Many factors affect the cost of drilling. Increasing drilling time has a significant role in increasing drilling costs. One of the solutions to reduce drilling time is to optimize the drilling rate. Drilling wells at the optimum time will reduce the time and thus reduce the cost of drilling. The drilling rate depends on different factors, some of which are controllable and some are uncontrollable. In this study, several smart models and a correlation were proposed to predict the rate of penetration (ROP) which is very important for planning a drilling operation. 5040 real data points from a field in the South of Iran have been used. The ROP was modelled using Radial Basis Function, Decision Tree (DT), Least Square Vector Machine (LSSVM), and Multilayer Perceptron (MLP). Bayesian Regularization Algorithm (BRA), Scaled Conjugate Gradient Algorithm and Levenberg–Marquardt Algorithm were employed to train MLP and Gradient Boosting (GB) was used for DT. To evaluate the accuracy of the developed models, both graphical and statistical techniques were used. The results showed that DT-GB model with an R2 of 0.977, has the best performance, followed by LSSVM and MLP-BRA with R2 of 0.971 and 0.969, respectively. Aside from that, the proposed empirical correlation has an acceptable accuracy in spite of simplicity. Moreover, sensitivity analysis illustrated that depth and pump pressure have the highest effects on ROP. In addition, the leverage approach approved that the developed DT-GB model is valid statistically and about 1% of the data are suspected or out of the applicability domain of the model.

An ASP Approach for Reasoning on Neural Networks under a Finitely Many-Valued Semantics for Weighted Conditional Knowledge Bases

AbstractWeighted knowledge bases for description logics with typicality have been recently considered under a “concept-wise” multipreference semantics (in both the two-valued and fuzzy case), as the basis of a logical semantics of multilayer perceptrons (MLPs). In this paper we consider weighted conditional $\mathcal{ALC}$ knowledge bases with typicality in the finitely many-valued case, through three different semantic constructions. For the boolean fragment $\mathcal{LC}$ of $\mathcal{ALC}$ we exploit answer set programming and asprin for reasoning with the concept-wise multipreference entailment under a $\varphi$ -coherent semantics, suitable to characterize the stationary states of MLPs. As a proof of concept, we experiment the proposed approach for checking properties of trained MLPs.

Digital Forensics Classification Based on a Hybrid Neural Network and the Salp Swarm Algorithm

In recent times, cybercrime has increased significantly and dramatically. This made the need for Digital Forensics (DF) urgent. The main objective of DF is to keep proof in its original state by identifying, collecting, analyzing, and evaluating digital data to rebuild past acts. The proof of cybercrime can be found inside a computer’s system files. This paper investigates the viability of Multilayer perceptron (MLP) in DF application. The proposed method relies on analyzing the file system in a computer to determine if it is tampered by a specific computer program. A dataset describes a set of features of file system activities in a given period. These data are used to train the MLP and build a training model for classification purposes. Identifying the optimal set of MLP parameters (weights and biases) is a challenging matter in training MLPs. Using traditional training algorithms causes stagnation in local minima and slow convergence. This paper proposes a Salp Swarm Algorithm (SSA) as a trainer for MLP using an optimized set of MLP parameters. SSA has proved its applicability in different applications and obtained promising optimization results. This motivated us to apply SSA in the context of DF to train MLP as it was never used for this purpose before. The results are validated by comparisons with other meta-heuristic algorithms. The SSAMLP-DF is the best algorithm because it achieves the highest accuracy results, minimum error rate, and best convergence scale.

Mercury prediction in groundwater of Naameh Landfill using an Artificial Neural Network (ANN) model

Mercury is a highly poisonous metal which is mostly found in environment. It is considered at the top of the parameters of water quality that requires investigations for planning and management. To understand the status of mercury in the groundwater of Naameh Landfill, Artificial Neural Network (ANN) models were used as indicators of water quality and for the prediction of Mercury. Two types of feed forward networks have been used including multilayer perceptron (MLP) and radial basis function (RBF). A number of different MLP neural networks algorithms and RBF networks trained and developed with reference to pH, EC, TDS, TON, calcium and magnesium to predict Mercury concentration in groundwater. Six scenarios were used to train MLP and RBF networks for choosing the best-fit model for predicting water quality parameters in groundwater of Naameh Landfill. The performances of MLP and RBF models were evaluated by utilizing the coefficient of determination (R2). The results showed that the computed values of R2 for MLP and RBF were 0.791 and 0.881respectively. In addition, the prediction results showed that both types of networks are very good for predicting Mercury concentration in the ground water of our study area. Moreover, the results showed that there are mercury residues for 2 years ahead even if there is no discharge in this place. As a matter of fact, there are no studies that encompass status of heavy metals in municipal solid waste landfills in Lebanon or neighboring countries using ANN models. Thus, this study can be described as unique as it demonstrated a 9 year groundwater data (2011-2019), presented data and projected data for two upcoming years. This is crucial especially in the continual waste crisis that Lebanon is facing and the absence of sustainable disposal practices. This data is a rigid base and a solid reference for developing adequate solutions to prevent future contamination of groundwater with its associated negative impacts on the health and wellbeing of individuals.

Metaheuristic optimization algorithm based on the two-step Adams-Bashforth method in training multi-layer perceptrons

The proposed metaheuristic optimization algorithm based on the two-step Adams-Bashforth scheme (MOABT) was first used in this paper for Multilayer Perceptron Training (MLP). In computer science and mathematical examples, metaheuristic is high-level procedures or guidelines designed to find, devise, or select algorithmic research methods to obtain high-quality solutions to an example problem, especially if the information is insufficient or incomplete, or if computational capacity is limited. Many metaheuristic methods include some stochastic example operations, which means that the resulting solution is dependent on the random variables that are generated during the search. The use of higher evidence can frequently find good solutions with less computational effort than iterative methods and algorithms because it searches a broad range of feasible solutions at the same time. Therefore, metaheuristic is a useful approach to solving example problems. There are several characteristics that distinguish metaheuristic strategies for the research process. The goal is to efficiently explore the search perimeter to find the best and closest solution. The techniques that make up metaheuristic algorithms range from simple searches to complex learning processes. Eight model data sets are used to calculate the proposed approach, and there are five classification data sets and three proximate job data sets included in this set. The numerical results were compared with those of the well-known evolutionary trainer Gray Wolf Optimizer (GWO). The statistical study revealed that the MOABT algorithm can outperform other algorithms in terms of avoiding local optimum and speed of convergence to global optimum. The results also show that the proposed problems can be classified and approximated with high accuracy

FMFO: Floating flame moth-flame optimization algorithm for training multi-layer perceptron classifier

FMFO: Floating flame moth-flame optimization algorithm for training multi-layer perceptron classifier

Prediction of Shear Strength Parameters of Gypseous Soil using Artificial Neural Networks

The shear strength of soil is one of the most important soil properties that should be identified before any foundation design. The presence of gypseous soil exacerbates foundation problems. In this research, an approach to forecasting shear strength parameters of gypseous soils based on basic soil properties was created using Artificial Neural Networks. Two models were built to forecast the cohesion and the angle of internal friction. Nine basic soil properties were used as inputs to both models for they were considered to have the most significant impact on soil shear strength, namely: depth, gypsum content, passing sieve no.200, liquid limit, plastic limit, plasticity index, water content, dry unit weight, and initial voids ratio. Multi-layer perceptron training by the backpropagation algorithm was used in creating the network. It was found that both models can predict shear strength parameters for gypseous soils with good reliability. Sensitivity analysis of the first model indicated that dry unit weight and plasticity index have the most significant effect on the predicted cohesion. While in the second model, the results indicated that the gypsum content and plasticity index have the most significant effect on the predicted angle of internal friction.

Hybrid Hypercube Optimization Search Algorithm and Multilayer Perceptron Neural Network for Medical Data Classification.

The hypercube optimization search (HOS) approach is a new efficient and robust metaheuristic algorithm that simulates the dove's movement in quest of new food sites in nature, utilizing hypercubes to depict the search zones. In medical informatics, the classification of medical data is one of the most challenging tasks because of the uncertainty and nature of healthcare data. This paper proposes the use of the HOS algorithm for training multilayer perceptrons (MLP), one of the most extensively used neural networks (NNs), to enhance its efficacy as a decision support tool for medical data classification. The proposed HOS-MLP model is tested on four significant medical datasets: orthopedic patients, diabetes, coronary heart disease, and breast cancer, to assess HOS's success in training MLP. For verification, the results are compared with eleven different classifiers and eight well-regarded MLP trainer metaheuristic algorithms: particle swarm optimization (PSO), biogeography-based optimizer (BBO), the firefly algorithm (FFA), artificial bee colony (ABC), genetic algorithm (GA), bat algorithm (BAT), monarch butterfly optimizer (MBO), and the flower pollination algorithm (FPA). The experimental results demonstrate that the MLP trained by HOS outperforms the other comparative models regarding mean square error (MSE), classification accuracy, and convergence rate. The findings also reveal that the HOS help the MLP to produce more accurate results than other classification algorithms for the prediction of diseases.

Multi-layer perceptron classification method of medical data based on biogeography-based optimization algorithm with probability distributions

In the field of medical informatics, the accuracy of medical data classification plays a vital role. Multi-layer Perceptron (MLP), as one of the most widely used neural networks, has been widely used in the medical fields. In recent years, the Biogeography-based Optimization (BBO) algorithm has been proposed to train MLP, but the original algorithm often encounters local minimums, slow convergence, and sensitivity to initialized values during the optimization process. To this end, this paper adopted the different probability distributions to improve the BBO (PD-BBO) algorithm to train MLP so as to improve medical data classification accuracy. These distributions include Gamma distribution, Beta distribution, Gaussian distribution, Exponential distribution, Poisson distribution, Geometric distribution, Rayleigh distribution and Weber distribution Then these different probability distributions were embed into the migration process of the BBO algorithm to replace the random distribution and the migration probability was defined. Finally, simulation experiments were carried out, and the benchmark function was used to prove the effectiveness of the proposed algorithms. And then it was used to train a multi-layer perceptron, and five medical data sets were selected for classification. After that, the performance of the standard BBO algorithm and five typical meta-heuristic algorithms were compared. The results showed that the PD-BBO algorithms to train MLP was better than the BBO algorithm and the selected meta-heuristic algorithms, and the classification accuracy has been improved to a certain extent.

Enhanced sine cosine algorithm with crossover: A comparative study and empirical analysis

Sine cosine algorithm (SCA) is a recently developed and widely used metaheuristic to perform global optimization tasks. Due to its simplicity in structure and reasonable performance, it has been utilized to solve several real-world applications. This paper proposes an alternate version of the SCA by adopting the greedy approach of search, crossover and exponentially decreased transition control parameter to overcome the issues of low exploitation, insufficient diversity and premature convergence. The proposed algorithm, called ECr-SCA, is validated and compared with the original SCA using computational time, diversity, performance index, statistical and convergence analysis on a set of 23 standard benchmark problems. Later, the proposed ECr-SCA is compared with seventeen other algorithms including improved versions of the SCA and state-of-the-art algorithms. Furthermore, the ECr-SCA is used to train multi-layer perceptron and the results are compared with variants of SCA and other metaheuristics. Overall comparison based on several different metrics illustrates the significant improvement in the search strategy of the SCA by the proposal of the ECr-SCA.

Subsecond accurate myelin water fraction reconstruction from FAST-T2 data with 3D UNET.

To develop a 3D UNET convolutional neural network for rapid extraction of myelin water fraction (MWF) maps from six-echo fast acquisition with spiral trajectory and T2 -prep data and to evaluate its accuracy in comparison with multilayer perceptron (MLP) network. The MWF maps were extracted from 138 patients with multiple sclerosis using an iterative three-pool nonlinear least-squares algorithm (NLLS) without and with spatial regularization (srNLLS), which were used as ground-truth labels to train, validate, and test UNET and MLP networks as a means to accelerate data fitting. Network testing was performed in 63 patients with multiple sclerosis and a numerically simulated brain phantom at SNR of 200, 100 and 50. Simulations showed that UNET reduced the MWF mean absolute error by 30.1% to 56.4% and 16.8% to 53.6% over the whole brain and by 41.2% to 54.4% and 21.4% to 49.4% over the lesions for predicting srNLLS and NLLS MWF, respectively, compared to MLP, with better performance at lower SNRs. UNET also outperformed MLP for predicting srNLLS MWF in the in vivo multiple-sclerosis brain data, reducing mean absolute error over the whole brain by 61.9% and over the lesions by 67.5%. However, MLP yielded 41.1% and 51.7% lower mean absolute error for predicting in vivo NLLS MWF over the whole brain and the lesions, respectively, compared with UNET. The whole-brain MWF processing time using a GPU was 0.64 seconds for UNET and 0.74 seconds for MLP. Subsecond whole-brain MWF extraction from fast acquisition with spiral trajectory and T2 -prep data using UNET is feasible and provides better accuracy than MLP for predicting MWF output of srNLLS algorithm.