Neural Network Optimization Research Articles

Convolutional neural networks optimization using multi-objective particle swarm optimization algorithm

This paper presents a heuristic-based convolutional neural network architecture optimization scheme (HeuCNN). First, CNN architectures are encoded as particles in optimization space with two objectives test error and number of parameters. Then, the effectiveness of CNN particles is represented by a heuristic in optimization space. It is mathematically proven and experimentally shown that the proposed heuristic assigns the best values for CNNs with higher performances. Uniform and nonuniform distribution of CNN particles in optimization space are analyzed and behavior of CNNs moving towards higher test error and higher number of parameters are also mathematically proven. CNN architecture search is proposed based on particle swarm optimization (PSO) by new algorithms for particle difference and velocity computation as well as particle update rules. Limitations of HeuCNN are explained and primary insights are presented to address its challenges as future works. Evaluation of HeuCNN on 9 publicly available datasets and comparing with 38 constant and optimized CNN peer competitive models reveal that HeuCNN achieves the best test error and the lowest number of parameters simultaneously. Source codes, data and software associated with this research will be available online after the publication of the paper.

Optimization of renewable energy hybrid power plant in the state of Karnataka, India

This paper proposes a methodology on reliability and cost minimization of the electricity produced by Renewable energy sources. The approach is to use Hybrid Renewable Energy Systems, which involves integrating various renewable sources. The objective of this study is to address these challenges by HRES method in the state of Karnataka. In Karnataka, there has been large scale curtailment of renewable energy generation due to the various structural issues. This study would give a solution to reliability and cost related issues faced by this state. For the study a Mathematical Model has been developed using Feed Forward Neural Network (FNN) and Chemical Reaction Optimization algorithms. Chemical reaction Optimization model is chosen to remove biases of the Artificial Neural Network Model of FNN. This model will be used to simulate interaction between Hybrid Renewable Energy Systems. The result of the study will provide a power system with high & optimized renewable energy output in the state of Karnataka, India which will be reliable and cost effective.

Interior sound quality evaluation and forecasting of passenger vehicles based on hybrid optimization neural networks

The interior noise of vehicles directly affects the comfort of the occupants, necessitating precise evaluation and control. Existing research has focused on constructing mappings between objective parameters and subjective perceptions of noise, where back propagation neural networks (BPNNs) are widely used due to their strong nonlinear mapping capabilities. However, the selection of initial weights and thresholds can affect the predictive accuracy of BPNN. This study developed a BPNN model optimized by an intelligent algorithm for predicting the level of subjective annoyance of passengers during the movement. Initially, objective parameters of interior noise were obtained through acoustic signal processing techniques, and five parameters were selected for studying subjective annoyance through correlation analysis and two-tailed tests. Meanwhile, the actual subjective ratings of passengers on interior noise were captured for subsequent training of the model and testing of the results. Finally, the established sparrow search algorithm (SSA) and genetic algorithm (GA) optimized BPNN were used to predict subjective evaluations. The predictive accuracy and efficiency of the model were significantly improved upon validation, providing a viable alternative to traditional passenger vehicle noise assessment experiments and valuable references for future noise control and optimization efforts. The experimental results are consistent with the view that the neural network model optimized with a mixture of intelligent algorithms is closer to the passenger’s subjective annoyance level having higher accuracy and efficiency.

Application of multi-objective neural network algorithm in industrial polymerization reactors for reducing energy cost and maximising productivity

Optimization on an industrial scale is a complex task that involves fine-tuning the performance of large-scale systems and applications to make them more efficient and effective. This process can be challenging due to the increasing volume of work, growing system complexity, and the need to maintain optimal performance. Due to the significant power required for compression and the high costs of reactant materials, optimizing low-density polyethylene (LDPE) production to provide maximum productivity with a reduction of energy cost is required. However, it is not a simple process because the optimization problem of the LDPE tubular reactor consists of conflicting objective functions. Multi-objective neural network algorithm (MONNA) is a metaheuristic optimization method that provides a versatile and robust approach for solving complex, contradictory targets and diverse optimization problems that do not rely on specific mathematical properties of the problem. It is inspired by the structure and information-processing capabilities of biological neural networks. MONNA iteratively proposes solutions, evaluates its performance, and adjusts its approach based on feedback, which avoids complex mathematical formulations. In this work, we implement Multi-objective optimization neural network algorithm (MONNA) in LDPE tubular reactor for maximising productivity, conversion and minimising energy costs with three scenario of problem optimization, i.e. maximising productivity and reducing energy cost for the first problem (P1); increasing conversion and reducing energy costs for the second problem (P2); and increasing productivity and reducing by-products for the third problem (P3). The results show that the highest productivity, highest conversion, and lowest energy are 545.1 mil. RM/year, 0.314, and 0.672 mil. RM/year. The extreme points in the Pareto Front (PF) for various bi-objective situations provide practitioners with helpful information for selecting the best trade-off for the operational strategy. According to their preferences, decision-makers can use the resulting Pareto to decide on the most acceptable alternative. The decision variable plots show that both initiators in the reacting zone highly affected the optimal solution with the opposite action.

An improved dual shear unified strength model (IDSUSM) considering strain softening effect

This study proposes an improved dual shear unified strength model by introducing the plastic internal variable which reflects the collective effects of strain softening, intermediate principal stress and unequal strength under tension and compression. The improved model is then simplified into simple forms for typical stress states, including uniaxial tension and compression, plane stress pure shear and tri-axial stress states. The smooth method and conjugate gradient method are utilized to facilitate its numerical implementation, avoiding numerical singularity and non-convergence in the solution process. The physical meanings of the parameters are further clarified and their values for self-compacting concrete are determined from the results of triaxial compression tests through a combination of direct determination, equation solution and back propagation (BP) neural network optimization. Validated against the test results, the improved model gives a more accurate prediction than the traditional dual shear unified strength model and Mohr-Coulomb model, in terms of both the overall trend and representative values. Validation results show that the improved model is applicable to materials for which the compressive strength is greater than the tensile strength and the tensile strength is greater than the shear strength.

Seismic vulnerability assessment of buildings using rapid visual screening for hybrid UMNN-RFO approach

This chapter proposes a hybrid approach for seismic vulnerability assessment of buildings using rapid visual screening (RVS). The Unconstrained Monotonic Neural Network (UMNN) and Red Fox Optimisation (RFO) are combined to create the hybrid technique, thus, the term ‘UMNN-RFO approach’. The UMNN algorithm predicts the RVS score and the RFO algorithm optimizes the RVS score. The RVS methods via Google Maps are being used to examine the Andaman and Nicobar Islands because it is one of the main earthquake prone regions in the world. A score was given after a screening that looked at Reinforced Concrete moment-resisting frames, load-bearing structures, steel frames (SF) and wooden frames (WF). By then, the MATLAB platform will have the proposed model implemented. The proposed approach outperforms all current techniques, including Granular Computing Artificial Neural Network (GrC-ANN), Particle Swarm Optimization-Back Propagation Neural Network (PSO-BPNN) and Genetic Algorithm-Artificial Neural Network (GA-ANN). The proposed method’s accuracy is 90%, determination coefficient is 0.91 and the Root Mean Square error is 0.27%. From the result, the proposed method concluded that older structures are more likely to sustain greater damage shortly than more recently built ones based on the score.

Optimal design of laminated thermoelectric cylindrical device based on neural network and thermo-electric-mechanical coupled model

The laminated thermoelectric cylindrical shell is favorable for solar thermal energy harvesting, but the thermo-electric-mechanical coupling field is relatively complex. An optimization design model of thermoelectric structure is required to promote its practical application. In this paper, based on variable separation method, thermo-electric-mechanical coupled model of laminated thermoelectric cylindrical device under thermal shock is constructed. Transient temperature field, thermal stress and power generation efficiency are obtained. Numerical results show that both power generation efficiency and thermal stress are affected by material parameters simultaneously. In order to analyze efficiency and structural reliability comprehensively, a fitness function is introduced in this paper. Based on the data which are collected from thermoelectric coupled model, a neural network is established and trained, which can predict and evaluate the performance of thermoelectric cylindrical shell based on the fitness function. The optimal material parameters for thermoelectric shells with high efficiency and low thermal stress are obtained. Combining neural network technique and the theoretical model developed in this paper, the computation time of numerical results can be reduced from 50 to 3 min, and this neural network optimization method has higher computational efficiency. These research results will provide guidance for the optimization design of cylindrical thermoelectric device.

Uncertain seismic response and robustness analysis of isolated continuous girder bridges

When calculating the seismic response of structures, conventional methods often neglect uncertainty and suffer from excessive computation. Although robustness metrics have been effective in evaluating structural safety, they are not suitable for isolated continuous girder bridges (ICGB). To address this issue, this paper introduces an uncertain seismic response and robustness assessment method tailored to ICGB: Firstly, the method considers uncertainty in material parameters and ground motion by constructing a data set using Monte Carlo simulation. Secondly, a genetic algorithm for parallel optimisation of radial basis function neural network (GAPO-RBFNN) is trained to predict the seismic responses of the bridges. Thirdly, a robustness metric specifically designed for bridges is derived and applied to bridge. Finally, the proposed robustness metric is validated on lead rubber bearing (LRB) and shape memory alloy-lead rubber bearing (SMA-LRB) bridges. Results indicate that the GAPO-RBFNN model accurately approximates the mapping relationship between structural parameters and seismic response, with an average error of only 0.32% and a 70% reduction in computation time compared with traditional methods. Moreover, the new robustness metric overcomes the limitations of the dismantled member method and is applicable to bridge. The robustness of the SMA-LRB-ICGB, strengthened by SMA, is improved, providing evidence of the effectiveness of the proposed robustness metric.

Assessment of machine learning models trained by molecular dynamics simulations results for inferring ethanol adsorption on an aluminium surface

Molecular dynamics (MD) simulations can reduce our need for experimental tests and provide detailed insight into the chemical reactions and binding kinetics. There are two challenges while dealing with MD simulations: one is the time and length scale limitations, and the latter is efficiently processing the massive amount of data resulting from the MD simulations and generating the proper reaction rates. In this work, we evaluated the use of regression machine learning (ML) methods to solve these two challenges by developing a framework for ethanol adsorption on an Aluminium (Al) slab. This framework comprises three main stages: first, an all-atom molecular dynamics model; second, ML regression models; and third, validation and testing. In stage one, the adsorption of ethanol molecules on the Al surface for various temperatures, velocities and concentrations is simulated using the large-scale atomic/molecular massively parallel simulator (LAMMPS) and ReaxFF. The outcome of stage one is utilised for training, testing, and validating the predictive models in stages two and three. We developed and evaluated 28 different ML models for predicting the number of adsorbed molecules over time, including linear regression, support vector machine (SVM), decision trees, ensemble, Gaussian process regression (GPR), neural network (NN) and Bayesian hyper-parameter optimisation models. Based on the results, the Bayesian-based GPR showed the highest accuracy and the lowest training time. The developed model can predict the number of adsorbed molecules for new cases within seconds, while MD simulations take a few weeks. This adsorption rate can then be used in macroscale simulations to tackle the time and length scale limitations. The proposed numerical framework has the potential to be generalised and, therefore, contribute to future low-cost binding reaction estimations, providing a valuable tool for industry and experimentalists.

Application and optimization of residual connection neural network in spacecraft thermal design

In thermal analysis modeling, the finite element method (FEM) is commonly used; however, it incurs high computational costs and complicates the global optimization of thermal parameters. To address these challenges, developing simplified surrogate models is crucial for enhancing analysis efficiency. Yet, constructing such models demands exceptional predictive accuracy, making conventional parameter adjustment methods inadequate for design needs. This paper introduces a novel Residual connection Neural Network model, called Res-NN, designed to approximate the CMOS finite element model. By employing residual connections, the Res-NN model significantly reduces fitting errors between networks, achieving a predictive accuracy of 94.6 %, which is 6.6 % higher than that of comparable Multi-Layer Perceptron (MLP) models. Moreover, Res-NN is over 100 times faster than traditional FEM in prediction speed, effectively circumventing the difficulties associated with parameter adjustments. To optimize the temperature fluctuations of the CMOS model, we utilized the Res-NN model as an iterative object within an optimization algorithm. Through experimental comparisons, we identified the PSO algorithm as the most effective option. The PSO optimization results demonstrated a chip temperature difference of 0.045 °C, with a simulation error of only 0.0649 °C, meeting design specifications and achieving a 61.7 % reduction in temperature variance compared to traditional thermal design method. This validates the superiority of the entire optimization process.

Thermal performance analysis and optimization of a heat pipe-based electronic thermal management system using experimental data-driven neuro-genetic technique

Heat pipes are highly efficient heat transfer devices and are used widely to dissipate heat generated by electronic components, thereby maintaining their operating temperatures and preventing overheating. The present study deals with the thermal performance prediction of a cylindrical heat pipe (length 380 mm) based electronic thermal management system using steady-state experimental investigation. The experiments are conducted under varying operational conditions, including heat inputs (10–50 W), condenser cooling water inlet temperatures (288.15, 293.15, and 298.15 K), and flow rates (10, 25, and 40 LPH). The study shows that the evaporator temperature of the heat pipe consistently rises with the increase in heat inputs for all tested conditions. The lowest evaporator temperature (311.42 K) is noticed for the cooling water inlet temperature of 288.15 K across all heat inputs. Notably, at 10 LPH, the heat pipe supplied with cooling water at 288.15 K exhibits reductions of up to 1.47 % and 2.29 % compared to 293.15 K and 298.15 K, respectively. Similar trends are also observed at 25 LPH and 40 LPH. The heat pipe’s thermal resistance is lowest (0.95 K/W) at a cooling water inlet temperature of 298.15 K. At 10 LPH, there are reductions of 10.19 % and 5.62 % compared to 288.15 K and 293.15 K, and at 40 LPH, reductions are 15.66 % and 7.89 %. The heat pipe’s evaporator heat transfer coefficient also plays a significant role and is maximum for the cooling water inlet temperature of 288.15 K at a flow rate of 10 LPH. To understand the complex behavior of the heat pipe and for better prediction of its thermal performance, a neuro-genetic (experimental data-driven artificial neural network and genetic algorithm) optimization technique is considered in the study. This predicts the optimal combination of operating parameters (heat input, cooling water inlet temperature, and flow rate) to minimize the heat pipe’s evaporator wall temperature which is found to be 312.54 K at a heat input of 10 W, flow rate of 10 LPH, and cooling water inlet temperature of 289.14 K. These input combinations are further validated through the experiments and the error in the heat pipe’s evaporator temperature is found to be 0.66 % only. Overall, this neuro-genetic technique underscores the importance of optimizing operational parameters for enhanced heat pipe performance and offers valuable insights for electronic cooling systems.

NEURAL NETWORKS AND DEEP LEARNING: ENHANCING AI THROUGH NEURAL NETWORK OPTIMIZATION

Neural networks and deep learning have profoundly impacted artificial intelligence (AI), driving advancements across numerous applications. However, optimizing these networks remains a critical challenge, necessitating sophisticated techniques and methodologies. This article explores the state-of-the-art in neural network optimization, delving into advanced gradient descent variants, regularization methods, learning rate schedulers, batch normalization, and cutting-edge architectures. We discuss their theoretical underpinnings, implementation complexities, and empirical results, providing insights into how these optimization strategies contribute to the development of high-performance AI systems. Case studies in image classification and natural language processing illustrate practical applications and outcomes. The article concludes with an examination of current challenges and future directions in neural network optimization, emphasizing the need for scalable, interpretable, and robust solutions.

Parallel hyperparameter optimization of spiking neural networks

Hyperparameter optimization of spiking neural networks (SNNs) is a difficult task which has not yet been deeply investigated in the literature. In this work, we designed a scalable constrained Bayesian based optimization algorithm that prevents sampling in non-spiking areas of an efficient high dimensional search space. These search spaces contain infeasible solutions that output no or only a few spikes during the training or testing phases, we call such a mode a “silent network”. Finding them is difficult, as many hyperparameters are highly correlated to the architecture and to the dataset. We leverage silent networks by designing a spike-based early stopping criterion to accelerate the optimization process of SNNs trained by spike timing dependent plasticity and surrogate gradient. We parallelized the optimization algorithm asynchronously, and ran large-scale experiments on heterogeneous multi-GPU Petascale architecture. Results show that by considering silent networks, we can design more flexible high-dimensional search spaces while maintaining a good efficacy. The optimization algorithm was able to focus on networks with high performances by preventing costly and worthless computation of silent networks.

Optimization control of wastewater treatment based on neural network and multi-objective optimization algorithm

In wastewater treatment, energy consumption and effluent quality are conflicting objectives. Improving effluent quality while reducing energy consumption has become a pressing issue. This paper proposes a dynamic optimal control method for the wastewater treatment process, utilizing neural networks and multi-objective dragonfly algorithm. A regression prediction model, based on a Self-Attention mechanism of a multi-scale convolutional neural network-bidirectional gated recurrent unit, is used to establish a soft measurement model for effluent quality and energy consumption, serving as the optimization objective function. The multi-objective dragonfly algorithm then optimizes this objective to determine the optimal set values for control variables. A Linear Active Disturbance Rejection Control, based on a scalable bandwidth linear extended state observer, is designed to track these optimal set values. The proposed method is tested using Benchmark Simulation Model No. 1 (BSM1). Results demonstrate that the method effectively obtains and tracks optimal set values of control variables, ensuring effluent quality meets standards and significantly reducing energy consumption compared to four other methods.

Parameter optimisation of marine four-stroke engine EGR systems using RSM and NSGA-III

ABSTRACT To reduce NOx emission of marine engines, an engine parameter optimisation method combining response surface method (RSM), Bayesian neural network and non-dominated sequence genetic algorithm (NSGA-III) is proposed. First, a simulation model of a marine four-stroke engine with an external exhaust gas recirculation (EGR) system was established and validated in AVL-BOOST software. Then, the five parameters of EGR valve opening, ignition advance angle, compression ratio (CR), intake valve opening (IVO) and exhaust valve closing (EVC) are selected as the decision parameters, and the three parameters of engine power, brake specific fuel consumption (BSFC) and NOx emission are taken as the target parameters to be optimised. In Design-Expert software, the response surface model is established by using the experimental design approach of central composite design (CCD), and the impact of decision parameters on target parameters are discussed by using response surface analysis. NSGA-III was used for the multi-objective optimisation of parameters, and the accuracy of NSGA-III was improved by the Bayesian neural network. The optimal solution is selected and verified by the technique for order of preference by similarity to the ideal solution (TOPSIS) approach. The results show that when the EGR valve opening is 47.5%, ignition advance angle is −18.99°CA, CR is 14.995, IVO is 15.27°CA and EVC is 7.68°CA, compared with the standard setting, the optimised power is increased by 4.17%, BSFC is reduced by 4.27%, and NOx emissions are reduced by 12.37%. Therefore, the combination of the Bayesian neural network and NSGA-III multi-objective optimisation can improve engine performance while reducing fuel consumption and NOx emissions.

Experimental Research and Improved Neural Network Optimization Based on the Ocean Thermal Energy Conversion Experimental Platform

With the progress of research on ocean thermal energy conversion, the stabI have checked and revised all. le operation of ocean thermal energy conversion experiments has become a problem that cannot be ignored. The control foundation for stable operation is the accurate prediction of operational performance. In order to achieve accurate prediction and optimization of the performance of the ocean thermal energy conversion experimental platform, this article analyzes the experimental parameters of the turbine based on the basic experimental data obtained from the 50 kW OTEC experimental platform. Through the selection and training of experimental data, a GA-BP-OTE (GBO) model that can automatically select the number of hidden layer nodes was established using seven input parameters. Bayesian optimization was used to complete the optimization of hyperparameters, greatly reducing the training time of the surrogate model. Analyzing the prediction results of the GBO model, it is concluded that the GBO model has better prediction accuracy and has a very low prediction error in the prediction of small temperature changes in ocean thermal energy, proving the progressiveness of the model proposed in this article. The dual-objective optimization problem of turbine grid-connected power and isentropic efficiency is solved. The results show that the change in isentropic efficiency of the permeable device is affected by the combined influence of the seven parameters selected in this study, with the mass flow rate of the working fluid having the greatest impact. The MAPE of the GBO model turbine grid-connected power is 0.24547%, the MAPE of the turbine isentropic efficiency is 0.04%, and the MAPE of the turbine speed is 0.33%. The Pareto-optimal solution for the turbine grid-connected power is 40.1792 kW, with an isentropic efficiency of 0.837439.

Droop control based energy management of distributed batteries using hybrid approach

In this manuscript proposes a hybrid SO-CCG-DLNN approach for a droop control based Battery Storage System (BSS). The proposed hybrid approach is combination of both Cascade Correlation Growing Deep Learning Neural Network (CCG-DLNN) and Snake Optimizer (SO). Commonly it is named as SO-CCG-DLNN approach. Micro grid (MG) the use of hybrid renewable energy systems is an advanced and useful technology for the growth of sustainability that is environmentally beneficial. An MG mainly consists of Photovoltaic (PV), wind, and BSS. The primary goal of this study is to control the State of Charge (SoC) and improve the power efficiency of the battery. The droop manages balance and electricity from the batteries provided in the DC MG. The sources of renewable energy projection for electricity consumption and the uncertainty were combined. Obtain the optimal energy management solutions by using SO optimizer and CCG-DLNN to predict the SoC level and manage how the dispersed batteries are charged and discharged. The proposed model is implemented in MATLAB/Simulink platform and contrasted with several existing methods like Wild Horse Optimizer (WHO), Heap-Based Optimizer (HBO) and Particle Swam Algorithm (PSO).

Maximizing solar energy efficiency with efficient interleaved boost converter with Three-Level neutral point clamped inverter for Grid-Connected photovoltaic using hybrid approach

The efficient interleaved boost converter (IBC) combined with the 3-level neutral point clamped (NPC) inverter for grid-connected photovoltaic systems (GCPVS) maximizes solar energy efficiency are presented. The proposed hybrid technique implies the combination of Double Attention Convolutional Neural Network (DACNN) and Starling Murmuration Optimization (SMO) algorithm and is usually referred as DACNN-SMO technique. Enhancing power quality at the Point of Common Coupling (PCC) while utilizing the rated capacity of the inverter into consideration is the main goal of the proposed approach. The grid-connected photovoltaic system additionally includes an efficient three-level NPC inverter and interleaved boost converter to decrease DC link voltage oscillation. In addition to preventing overheating, the inverter current controls reactive power adjustment, active power injection, and current harmonic filtering. Utilizing the SMO technique, By employing the SMO technique, the system effectively regulates output voltage, ensuring that the inverter output voltage. The DACNN enhances the system’s ability to monitor and diagnose faults. Using MATLAB, the proposed topology is implemented, and the results are contrasted with those of other existing techniques. The existing methods such as Heap Based Optimizer (HBO), Circle Search Algorithm (CSA) and Sparrow Search Algorithm (SSA) attains a higher THD of 3.7%, 4.7% and 5.3% respectively. The proposed method DACNN-SMO attains a THD of 2.5%. The proposed technique displays the efficiency is 99.86%. When compared to existing strategies, the proposed technique displays lower THD and high efficiency.

The use of artificial neural networks in studying the progression of glaucoma

In ophthalmology, artificial intelligence methods show great promise due to their potential to enhance clinical observations with predictive capabilities and support physicians in diagnosing and treating patients. This paper focuses on modelling glaucoma evolution because it requires early diagnosis, individualized treatment, and lifelong monitoring. Glaucoma is a chronic, progressive, irreversible, multifactorial optic neuropathy that primarily affects elderly individuals. It is important to emphasize that the processed data are taken from medical records, unlike other studies in the literature that rely on image acquisition and processing. Although more challenging to handle, this approach has the advantage of including a wide range of parameters in large numbers, which can highlight their potential influence. Artificial neural networks are used to study glaucoma progression, designed through successive trials for near-optimal configurations using the NeuroSolutions and PyTorch frameworks. Furthermore, different problems are formulated to demonstrate the influence of various structural and functional parameters on the study of glaucoma progression. Optimal neural networks were obtained using a program written in Python using the PyTorch deep learning framework. For various tasks, very small errors in training and validation, under 5%, were obtained. It has been demonstrated that very good results can be achieved, making them credible and useful for medical practice.

An integrated convolutional neural network prediction framework for in situ shale oil content based on conventional logging data

The quantification of total organic carbon (TOC) and the free hydrocarbon content (S 1 ) is crucial for evaluating the shale oil generation and bearing properties of source rocks. This study aimed to enhance the accuracy of TOC and S 1 quantification in shale oil evaluations. The scope encompassed the development of a novel deep learning framework to overcome the limitations of traditional physical and machine learning or deep learning methods. This paper proposes an integrated swarm optimization algorithm–convolutional neural network/machine learning framework. This framework uses a swarm optimization algorithm for the hyperparameter optimization of a convolutional neural network or machine learning framework, utilizing experimental data from core samples preserved in liquid nitrogen alongside well logging data. The application of the proposed framework to the H11 well in the Subei Basin, China, using 110 core samples, demonstrated a superior performance. The results validate the framework's effectiveness in predicting the TOC and S 1 contents at various depths. The proposed framework stands out for its convenient methodology, wide application range and high precision in prediction. These attributes contribute significantly to the field of petroleum engineering and development, offering a novel approach that promises to enhance both the efficiency and accuracy of well logging evaluation.