Simulated Datasets Research Articles

DeepDive: estimating global biodiversity patterns through time using deep learning.

Understanding how biodiversity has changed through time is a central goal of evolutionary biology. However, estimates of past biodiversity are challenged by the inherent incompleteness of the fossil record, even when state-of-the-art statistical methods are applied to adjust estimates while correcting for sampling biases. Here we develop an approach based on stochastic simulations of biodiversity and a deep learning model to infer richness at global or regional scales through time while incorporating spatial, temporal and taxonomic sampling variation. Our method outperforms alternative approaches across simulated datasets, especially at large spatial scales, providing robust palaeodiversity estimates under a wide range of preservation scenarios. We apply our method on two empirical datasets of different taxonomic and temporal scope: the Permian-Triassic record of marine animals and the Cenozoic evolution of proboscideans. Our estimates provide a revised quantitative assessment of two mass extinctions in the marine record and reveal rapid diversification of proboscideans following their expansion out of Africa and a >70% diversity drop in the Pleistocene.

Application and Comparison of Machine Learning and Database-Based Methods in Taxonomic Classification of High-Throughput Sequencing Data.

The advent of high-throughput sequencing technologies has not only revolutionized the field of bioinformatics but has also heightened the demand for efficient taxonomic classification. Despite technological advancements, efficiently processing and analysing the deluge of sequencing data for precise taxonomic classification remains a formidable challenge. Existing classification approaches primarily fall into two categories, database-based methods and machine learning methods, each presenting its own set of challenges and advantages. On this basis, the aim of our study was to conduct a comparative analysis between these two methods while also investigating the merits of integrating multiple database-based methods. Through an in-depth comparative study, we evaluated the performance of both methodological categories in taxonomic classification by utilizing simulated datasets. Our analysis revealed that database-based methods excel in classification accuracy when backed by a rich and comprehensive reference database. Conversely, while machine learning methods show superior performance in scenarios where reference sequences are sparse or lacking, they generally show inferior performance compared to database methods under most conditions. Moreover, our study confirms that integrating multiple database-based methods does, in fact, enhance classification accuracy. These findings shed new light on the taxonomic classification of high-throughput sequencing data and bear substantial implications for the future development of computational biology. For those interested in further exploring our methods, the source code of this study is publicly available on https://github.com/LoadStar822/Genome-Classifier-Performance-Evaluator. Additionally, a dedicated webpage showcasing our collected database, datasets, and various classification software can be found at: http://lab.malab.cn/∼tqz/project/taxonomic/.

A two‐dimensional, reach‐scale implementation of space‐time image velocimetry (STIV) and comparison to particle image velocimetry (PIV)

AbstractImage‐based algorithms have become a powerful tool for estimating flow velocities in rivers. In this study, we generalize the space‐time image velocimetry (STIV) framework for reach‐scale application rather than along a cross section. The new algorithm provides information on both the magnitude and orientation of velocity vectors, and we refer to the algorithm as two‐dimensional STIV, or 2D‐STIV. The workflow involves setting up a grid, using centreline tangent vectors as initial estimates of flow direction, and then extracting space‐time images (STIs) along search lines radiating from each grid node. The autocorrelation function is used to infer the inclination of streak lines present in STIs, which represents the advection of water surface features. Information on flow direction is obtained by evaluating various candidate search lines and identifying that which yields the highest velocity. This search can be performed exhaustively or via optimization. We applied the new 2D‐STIV algorithm to three test cases, one simulated data set and two natural channels, and compared image‐derived velocities to modelled or measured values. We also applied two established particle image velocimetry (PIV) algorithms to the same data sets. 2D‐STIV performed as well as the two PIV algorithms for simulated images. For a natural river with distinct water surface features, 2D‐STIV was effective for much of the channel but also led to a more patchy, irregular velocity field than the two PIV algorithms. For a site lacking obvious surface features, exhaustive 2D‐STIV led to velocity estimates uncorrelated with field data while the optimization‐based version produced erratic flow directions. 2D‐STIV also required greater image sequence durations, higher frame rates, and generally longer computational run times. Overall, ensemble PIV was the most reliable algorithm.

Rapid in-situ calibration of computational micro-spectrometer with few-shot meta-learning

Computational micro-spectrometers comprised of detector arrays and encoding structure arrays, such as on-chip Fabry-Perot (FP) cavity filters, have great potential in many in-situ applications owing to their compact size and snapshot imaging ability. Given manufacturing deviation and environmental influence are inevitable, easy and effective calibration for spectrometer is necessary, especially for in-situ applications. Currently calibration strategies based on iterative algorithms or neural networks require accurate measurements of pixel-level (spectral) encoding functions through monochromator or large amounts of standard samples. These procedures are time-consuming and expensive, thereby impeding in-situ applications. Meta-learning algorithms with few-shot learning ability can address this challenge by incorporating the prior knowledge in the simulated dataset. In this work, we propose a meta-learning algorithm free of measuring encoding function or large amounts of standard samples to calibrate a micro-spectrometer with manufacturing deviation effectively. Our micro-spectrometer comprises 16 types of FP filters covering a wavelength range of 550-720 nm. The center wavelength of each filter type deviates from the design up to 6 nm. After calibration with 15 different color data, the average reconstruction error on the test dataset decreased from 7.2 × 10 − 3 to 1.2 × 10 − 3, and further decreased to 9.4 × 10 − 4 when the calibration data increased to 24. The performance is comparable to algorithms trained with measured encoding function both in reconstruction error and generalization ability. We estimated that the cost of in-situ calibration through reflectance measurements of color chart decreased to one percent of the cost through monochromator measurements. By exploiting prior deviation information in simulation data with meta-learning, the efficiency and cost of calibration are significantly improved, thereby facilitating the large-scale production and in-situ application of micro-spectrometers.

An improved Baldwin–Lomax algebraic wall model for high-speed canonical turbulent boundary layers using established scalings

In this work, we employ well-established relations for compressible turbulent mean flows, including the velocity transformation and algebraic temperature–velocity (TV) relation, to systematically improve the algebraic Baldwin–Lomax (BL) wall model for high-speed zero-pressure-gradient air boundary layers. Any new functions or coefficients fitted by ourselves are avoided. Twelve published direct numerical simulation (DNS) datasets are employed for a priori inspiration and a posteriori examination, with Mach numbers up to 14 under adiabatic, cold and heated walls. The baseline BL model is the widely used one with semilocal scalings. Three targeted modifications are made. First, we employ a total-stress-based transformation (Griffin et al., Proc. Natl Acad. Sci. USA, vol. 118, issue 34, 2021, e2111144118) to the inner-layer eddy viscosity for improved scaling up to the logarithmic region. Second, we utilize the van Driest transformation in the outer layer based on the compressible defect velocity scaling. Third, considering the difficulty in modelling the rapidly varying and singular turbulent Prandtl number near the temperature peak in cold-wall cases, we design a two-layer strategy and use the TV relation to formulate the inner-layer temperature. Numerical results prove that the modifications take effect as designed. The prediction accuracy for mean streamwise velocity is notably improved for diabatic cases, especially in the logarithmic region. Moreover, a significant improvement in mean temperature is realized for both adiabatic and diabatic cases. The mean relative errors of temperature to DNS for all cases are down to 0.4 % in the logarithmic wall-normal coordinate and 3.4 % in the outer coordinate, around one-third of those in the baseline model.

Imaging through thick scattering media based on envelope-informed learning with a simulated training dataset

Computational imaging faces significant challenges in dealing with multiple scattering through thick complex media. While deep learning has addressed some ill-posed problems in scattering imaging, its practical application is limited by the acquisition of the training dataset. In this study, the Gaussian-distributed envelope of the speckle image is employed to simulate the point spread function (PSF), and the training dataset is obtained by the convolution of the handwritten digits with the PSF. This approach reduces the requirement of time and conditions for constructing the training dataset and enables a neural network trained on this dataset to reconstruct objects obscured by an unknown scattering medium in real experiments. The quality of reconstructed objects is negatively correlated with the thickness of the scattering medium. Our proposed method provides a new way, to the best of our knowledge, to apply deep learning in scattering imaging by reducing the time needed for constructing the training dataset.

Hunger Games Search for the elucidation of gravity anomalies with application to geothermal energy investigations and volcanic activity studies

Abstract Recent metaheuristic approaches are extensively and intensively being implemented to the interpretation of gravity anomalies due to their superior advantages. We emphasize the application of Hunger Games Search (HGS), a newly established metaheuristic inspired by hunger-driven instincts and behavioral choices of animals, to elucidate gravity data for geothermal energy exploration and volcanic activity study. After recognizing the modal features of the objective function tailored and tuning the algorithm control parameters involved, HGS has been trial-tested on simulated data sets of different scenarios and finally experienced in two field cases from India and Japan. Notably, a second moving average strategy has been successfully integrated into the objective function to eradicate the regional component from observed responses. Post-inversion uncertainty appraisal tests have been further implemented to comprehend the reliability of solutions obtained. The solutions retrieved by HGS have been unbiasedly compared in terms of convergence rate, accuracy, stability, and robustness with the solutions of the commonly used particle swarm optimization algorithm. Based on the results accessed, the theoretical and field cases presented could be recuperated more precisely, stably, robustly, and coherently with the available geophysical, geological, and borehole verification, as HGS is able to better explore the model space without compromising its capability to efficiently approach the global minimum. This novel global optimization method can thus be considered as a promising tool in geothermal energy investigations and the study of volcanic activities.

A simulation study to assess the impact missing values on the performance of different statistical methods for analysis of binary repeated measures data with an additional hierarchical structure

The primary objective of the study was to assess the impact of missing values on the analy- sis of binary repeated measures data with an additional hierarchical structure. One motivat- ing example for the present study was records of high somatic cell counts in milk samples obtained by approximately monthly sampling throughout the lactations of cows in dairy herds. Random effects models with autocorrelated (ρ = 1, 0.9 or 0.5) subject-level ran- dom effects were behind the simulated data. In general, the settings of the simulation were chosen to reflect a real somatic cell count dataset (scc40), except that the within-cow time series length was set to 8-time points for each cow. The estimation procedures consid- ered were: Ordinary Logistic Regression (OLR), Alternating Logistic Regression (ALR), Weighted Generalized Estimating Equations (WGEE), Penalized Quasi Likelihood (PQL), Maximum likelihood via numerical integration (ML) and Bayesian Markov chain Monte Carlo (MCMC). Multiple scenarios of simulated incomplete datasets were considered and include: a scenario corresponded to a combination of missingness patterns present in the scc40 dataset (scc40 scenario) The remaining scenarios involved only drop-outs, and corre- sponded to either moderate or high percentages of values either missing at random (MAR) or not missing at random (NMAR), respectively. In the scc40 scenario, all estimation procedures except OLR performed well and produced estimates with small relative bias (generally less than 5%) for levels of missingness that roughly corresponded to the scc40 data. In MAR missingness scenarios, some biases were found for ALR, WGEE and PQL procedures, whereas the likelihood-based procedures were largely unaffected by the miss- ing values. In NMAR scenarios, all procedures experienced similar and strong biases in the time coefficient; however, fixed effects estimates at the subject and cluster levels were relatively unaffected. Journal of Statistical Research 2023, Vol 57, No.1-2, pp.35-67

Inferring causal protein signalling networks from single‐cell data based on parallel discrete artificial bee colony algorithm

AbstractInferring causal protein signalling networks from human immune system cell data is a promising approach to unravel the underlying tissue signalling biology and dysfunction in diseased cells, which has attracted considerable attention within the bioinformatics field. Recently, Bayesian network (BN) techniques have gained significant popularity in inferring causal protein signalling networks from multiparameter single‐cell data. However, current BN methods may exhibit high computational complexity and ignore interactions among protein signalling molecules from different single cells. A novel BN method is presented for learning causal protein signalling networks based on parallel discrete artificial bee colony (PDABC), named PDABC. Specifically, PDABC is a score‐based BN method that utilises the parallel artificial bee colony to search for the global optimal causal protein signalling networks with the highest discrete K2 metric. The experimental results on several simulated datasets, as well as a previously published multi‐parameter fluorescence‐activated cell sorter dataset, indicate that PDABC surpasses the existing state‐of‐the‐art methods in terms of performance and computational efficiency.

Comparison of some Bayesian estimation methods for type-I generalized extreme value distribution with simulation

The Weibull distribution is considered one of the Type-I Generalized Extreme Value (GEV) distribution, and it plays a crucial role in modeling extreme events in various fields, such as hydrology, finance, and environmental sciences. Bayesian methods play a strong, decisive role in estimating the parameters of the GEV distribution due to their ability to incorporate prior knowledge and handle small sample sizes effectively. In this research, we compare several shrinkage Bayesian estimation methods based on the squared error and the linear exponential loss functions. They were adopted and compared by the Monte Carlo simulation method. The performance of these methods is assessed based on their accuracy and computational efficiency in estimating the scale parameter of the Weibull distribution. To evaluate their performance, we generate simulated datasets with different sample sizes and varying parameter values. A technique for pre-estimation shrinkage is suggested to enhance the precision of estimation. Simulation experiments proved that the Bayesian shrinkage estimator and shrinkage pre-estimation under the squared loss function method are better than the other methods because they give the least mean square error. Overall, our findings highlight the advantages of shrinkage Bayesian estimation methods for the proposed distribution. Researchers and practitioners in fields reliant on extreme value analysis can benefit from these findings when selecting appropriate Bayesian estimation techniques for modeling extreme events accurately and efficiently.

The all-pairs vitality-maximization (VIMAX) problem

AbstractTraditional network interdiction problems focus on removing vertices or edges from a network so as to disconnect or lengthen paths in the network; network diversion problems seek to remove vertices or edges to reroute flow through a designated critical vertex or edge. We introduce the all-pairs vitality maximization problem (VIMAX), in which vertex deletion attempts to maximize the amount of flow passing through a critical vertex, measured as the all-pairs vitality of the vertex. The assumption in this problem is that in a network for which the structure is known but the physical locations of vertices may not be known (e.g., a social network), locating a person or asset of interest might require the ability to detect a sufficient amount of flow (e.g., communications or financial transactions) passing through the corresponding vertex in the network. We formulate VIMAX as a mixed integer program, and show that it is NP-Hard. We compare the performance of the MIP and a simulated annealing heuristic on both real and simulated data sets and highlight the potential increase in vitality of key vertices that can be attained by subset removal. We also present graph theoretic results that can be used to narrow the set of vertices to consider for removal.

Visual Reinforcement Learning for Dynamic Object Detection

Abstract. Object detection is a widely studied task in computer vision. Current methods often focus on images captured from appropriate viewpoints. However, there is a large disparity between objects observed from different viewpoints in the real world. Dynamic Object Detection (DOD) method automatically adjusts the camera viewpoint in a visual scene to sequentially find optimal viewpoints. Currently, the DOD tasks are usually modeled as a sequential decision-making problem and solved using reinforcement learning methods. Existing approaches face challenges with sparse rewards and training instability. To tackle these issues, we proposed a single-step reward function and a lightweight network, respectively. The single-step reward function, which provides timely feedback, gives an efficient training process for DOD tasks. The lightweight network with few parameters can ensure the stability of the training process. To evaluate the effectiveness of our method, we developed a simulation dataset based on UE4, which consists of 1800 training images and 450 testing images. The dataset includes five object categories: vans, cars, trailers, box trucks and SUVs. Experiments demonstrate that our method outperforms SOTA object detectors on our simulation dataset. Specifically, the average precisions(APs) are improved from 89.1% to 96.0% when using the YOLOv8 object detector.

Three-Way Alignment Improves Multiple Sequence Alignment of Highly Diverged Sequences

The standard approach for constructing a phylogenetic tree from a set of sequences consists of two key stages. First, a multiple sequence alignment (MSA) of the sequences is computed. The aligned data are then used to reconstruct the phylogenetic tree. The accuracy of the resulting tree heavily relies on the quality of the MSA. The quality of the popularly used progressive sequence alignment depends on a guide tree, which determines the order of aligning sequences. Most MSA methods use pairwise comparisons to generate a distance matrix and reconstruct the guide tree. However, when dealing with highly diverged sequences, constructing a good guide tree is challenging. In this work, we propose an alternative approach using three-way dynamic programming alignment to generate the distance matrix and the guide tree. This three-way alignment incorporates information from additional sequences to compute evolutionary distances more accurately. Using simulated datasets on two symmetric and asymmetric trees, we compared MAFFT with its default guide tree with MAFFT with a guide tree produced using the three-way alignment. We found that (1) the three-way alignment can reconstruct better guide trees than those from the most accurate options of MAFFT, and (2) the better guide tree, on average, leads to more accurate phylogenetic reconstruction. However, the improvement over the L-INS-i option of MAFFT is small, attesting to the excellence of the alignment quality of MAFFT. Surprisingly, the two criteria for choosing the best MSA (phylogenetic accuracy and sum-of-pair score) conflict with each other.

Beta transformation of the Exponential-Gaussian distribution with its properties and applications

This study introduces a five-parameter continuous probability model named the Beta-Exponential-Gaussian distribution by extending the three-parameter Exponential-Gaussian distribution with the beta transformation method. The basic properties of the new distribution, including reliability measure, hazard function, survival function, moment, skewness, kurtosis, order statistics, and asymptotic behavior, are established. Using the acceptance-rejection algorithm, simulation studies are conducted. The new model is fitted to the simulated and real data sets, and its performance is reported. The Beta-Exponential-Gaussian distribution is found to be more flexible and has better performance in many aspects. It is suggested that the new distribution would be used in modeling data having skewness and bimodal distribution.

A Long-time-series Spatio-Temporal-Spectral Fusion Method via Multi-task Learning

Abstract. Due to the limitations of sensor hardware, clouds and fog, and data transmission limitations, it is difficult for the data obtained by spaceborne remote sensing imager to achieve high temporal, spatial and spectral resolution at the same time, which limits its application in long-time-series high-frequency monitoring. At present, there are several spatio-temporal-spectral algorithms that can realize the fusion of temporal, spatial and spectral resolution, but most of them are based on one to two discrete images, and the integrated fusion at the multi-dimensional level has not yet been realized. There is currently no research on the spatio-temporal-spectral fusion method based on LONG-TIME-SERIES multi-scene remote sensing data. Aiming at solving the bottleneck of spatio-tempora-spectral resolution of remote sensing data, this study proposes a new long-time-series spatio-temporal-spectral fusion method based on multi-task learning to realize the multi-dimensional optimization of multi-source remote sensing data resolutions. Experiments used simulated and real datasets, both of which contain 4 images of 10m ZY1-02D multispectral data, 7 images of 16m GF-6 multispectral data and 4 images of 30m ZY1-02D hyperspectral data, and obtained 7 images of 10m hyperspectral data. The experiments The results show that our method performs the best compared to other methods. This method can provide effective data support for applications based on long-time series remote sensing data.

Multi-Input data ASsembly for joint Analysis (MIASA): A framework for the joint analysis of disjoint sets of variables.

The joint analysis of two datasets [Formula: see text] and [Formula: see text] that describe the same phenomena (e.g. the cellular state), but measure disjoint sets of variables (e.g. mRNA vs. protein levels) is currently challenging. Traditional methods typically analyze single interaction patterns such as variance or covariance. However, problem-tailored external knowledge may contain multiple different information about the interaction between the measured variables. We introduce MIASA, a holistic framework for the joint analysis of multiple different variables. It consists of assembling multiple different information such as similarity vs. association, expressed in terms of interaction-scores or distances, for subsequent clustering/classification. In addition, our framework includes a novel qualitative Euclidean embedding method (qEE-Transition) which enables using Euclidean-distance/vector-based clustering/classification methods on datasets that have a non-Euclidean-based interaction structure. As an alternative to conventional optimization-based multidimensional scaling methods which are prone to uncertainties, our qEE-Transition generates a new vector representation for each element of the dataset union [Formula: see text] in a common Euclidean space while strictly preserving the original ordering of the assembled interaction-distances. To demonstrate our work, we applied the framework to three types of simulated datasets: samples from families of distributions, samples from correlated random variables, and time-courses of statistical moments for three different types of stochastic two-gene interaction models. We then compared different clustering methods with vs. without the qEE-Transition. For all examples, we found that the qEE-Transition followed by Ward clustering had superior performance compared to non-agglomerative clustering methods but had a varied performance against ultrametric-based agglomerative methods. We also tested the qEE-Transition followed by supervised and unsupervised machine learning methods and found promising results, however, more work is needed for optimal parametrization of these methods. As a future perspective, our framework points to the importance of more developments and validation of distance-distribution models aiming to capture multiple-complex interactions between different variables.

No-Reference Hyperspectral Image Quality Assessment via Ranking Feature Learning

In hyperspectral image (HSI) reconstruction tasks, due to the lack of ground truth in real imaging processes, models are usually trained and validated on simulation datasets and then tested on real measurements captured by real HSI imaging systems. However, due to the gap between the simulation imaging process and the real imaging process, the best model validated on the simulation dataset may fail on real measurements. To obtain the best model for the real-world task, it is crucial to design a suitable no-reference HSI quality assessment metric to reflect the reconstruction performance of different models. In this paper, we propose a novel no-reference HSI quality assessment metric via ranking feature learning (R-NHSIQA), which calculates the Wasserstein distance between the distribution of the deep features of the reconstructed HSIs and the benchmark distribution. Additionally, by introducing the spectral self-attention mechanism, we propose a Spectral Transformer (S-Transformer) to extract the spatial-spectral representative deep features of HSIs. Furthermore, to extract quality-sensitive deep features, we use quality ranking as a pre-training task to enhance the representation capability of the S-Transformer. Finally, we introduce the Wasserstein distance to measure the distance between the distribution of the deep features and the benchmark distribution, improving the assessment capacity of our method, even with non-overlapping distributions. The experimental results demonstrate that the proposed metric yields consistent results with multiple full-reference image quality assessment (FR-IQA) metrics, validating the idea that the proposed metric can serve as a substitute for FR-IQA metrics in real-world tasks.

A new sample-size planning approach for person-specific VAR(1) studies: Predictive accuracy analysis.

Researchers increasingly study short-term dynamic processes that evolve within single individuals using N = 1 studies. The processes of interest are typically captured by fitting a VAR(1) model to the resulting data. A crucial question is how to perform sample-size planning and thus decide on the number of measurement occasions that are needed. The most popular approach is to perform a power analysis, which focuses on detecting the effects of interest. We argue that performing sample-size planning based on out-of-sample predictive accuracy yields additional important information regarding potential overfitting of the model. Predictive accuracy quantifies how well the estimated VAR(1) model will allow predicting unseen data from the same individual. We propose a new simulation-based sample-size planning method called predictive accuracy analysis (PAA), and an associated Shiny app. This approach makes use of a novel predictive accuracy metric that accounts for the multivariate nature of the prediction problem. We showcase how the values of the different VAR(1) model parameters impact power and predictive accuracy-based sample-size recommendations using simulated data sets and real data applications. The range of recommended sample sizes is smaller for predictive accuracy analysis than for power analysis.

Range-Spread Target Detection Networks Using HRRPs

Range-spread target (RST) detection is an important issue for high-resolution radar (HRR). Traditional detectors relying on manually designed detection statistics have their performance limitations. Therefore, in this work, two deep learning-based detectors are proposed for RST detection using HRRPs, i.e., an NLS detector and DFCW detector. The NLS detector leverages domain knowledge from the traditional detector, treating the input HRRP as a low-level feature vector for target detection. An interpretable NLS module is designed to perform noise reduction for the input HRRP. The DFCW detector takes advantage of the extracted high-level feature map of the input HRRP to improve detection performance. It incorporates a feature cross-weighting module for element-wise feature weighting within the feature map, considering the channel and spatial information jointly. Additionally, a nonlinear accumulation module is proposed to replace the conventional noncoherent accumulation operation in the double-HRRP detection scenario. Considering the influence of the target spread characteristic on detector performance, signal sparseness is introduced as a measure and used to assist in generating two datasets, i.e., a simulated dataset and measured dataset incorporating real target echoes. Experiments based on the two datasets are conducted to confirm the contribution of the designed modules to detector performance. The effectiveness of the two proposed detectors is verified through performance comparison with traditional and deep learning-based detectors.

A physical simulation-machine learning model for optimal process schemes in laser-based directed energy deposition process

The major challenge faced is the definition of optimal process variables for rich quality of fabricated parts in laser-based directed energy deposition (DED-LB) processes. However, predicting the optimal process scheme using machine learning models is still challenging owing to the need for a large amount of training experimental data with high costs in DED-LB. In view of this, a physical simulation (PS)-machine learning (ML) model using a relatively small accurate data set of 31 simulation data for optimal process schemes is proposed in DED-LB process in this study. Firstly, a powder-scale high precision phenomenon model incorporating the mass transfer, phase transformations and heat transfer and using Lagrangian particle model to add mass is developed to depict the DED-LB process. Then, a Gaussian process regression (GPR) agent model is established to rapidly and accurately predict the geometry and dilution rate of deposition tracks under different manufacturing parameters based on the high precision simulation results. Finally, a PS-ML model for process parameter optimization is developed using the particle swarm algorithm (PSO), and the optimized parameters are experimentally validated. The results show that the developed powder-scale model and GPR model results are consistent with the Inconel 718 alloy experimental results. The proposed PS-ML model can increase the accuracy of the ML models even with a small simulation data set. None of PS-ML model optimization results has a relative error of more than 3% to the experimental results, and the dilution rate is reduced by up to 61.66% compared to the experimental design parameters without optimization. The proposed physical simulation-machine learning model in this study enables inexpensive and accurate optimization of DED-LB process parameters.