Optimizing Deep Learning for Computer-Aided Diagnosis of Lung Diseases: An Automated Method Combining Evolutionary Algorithm, Transfer Learning, and Model Compression

Recent developments in Computer Vision have presented novel opportunities to tackle complex healthcare issues,particularly in the field of lung disease diagnosis.One promising avenue involves the use of chest X-Rays,which are commonly utilized in radiology.To fully exploit their potential,researchers have suggested utilizing deep learning methods to construct computer-aided diagnostic systems.However,constructing and compressing these systems presents a significant challenge,as it relies heavily on the expertise of data scientists.To tackle this issue,we propose an automated approach that utilizes an evolutionary algorithm(EA)to optimize the design and compression of a convolutional neural network(CNN)for X-Ray image classification.Our approach accurately classifies radiography images and detects potential chest abnormalities and infections,including COVID-19.Furthermore,our approach incorporates transfer learning,where a pre-trainedCNNmodel on a vast dataset of chest X-Ray images is fine-tuned for the specific task of detecting COVID-19.This method can help reduce the amount of labeled data required for the task and enhance the overall performance of the model.We have validated our method via a series of experiments against state-of-the-art architectures.

A Deep-Learning and Transfer-Learning Hybrid Aerosol Retrieval Algorithm for FY4-AGRI:Development and Verification over Asia

The Advanced Geosynchronous Radiation Imager(AGRI)is a mission-critical instrument for the Fengyun series of satellites.AGRI acquires full-disk images every 15 min and views East Asia every 5 min through 14 spectral bands,enabling the detection of highly variable aerosol optical depth(AOD).Quantitative retrieval of AOD has hitherto been challenging,especially over land.In this study,an AOD retrieval algorithm is proposed that combines deep learning and transfer learning.The algorithm uses core concepts from both the Dark Target(DT)and Deep Blue(DB)algorithms to select features for the machinelearning(ML)algorithm,allowing for AOD retrieval at 550 nm over both dark and bright surfaces.The algorithm consists of two steps:①A baseline deep neural network(DNN)with skip connections is developed using 10 min Advanced Himawari Imager(AHI)AODs as the target variable,and②sunphotometer AODs from 89 ground-based stations are used to fine-tune the DNN parameters.Out-of-station validation shows that the retrieved AOD attains high accuracy,characterized by a coefficient of determination(R2)of 0.70,a mean bias error(MBE)of 0.03,and a percentage of data within the expected error(EE)of 70.7%.A sensitivity study reveals that the top-of-atmosphere reflectance at 650 and 470 nm,as well as the surface reflectance at 650 nm,are the two largest sources of uncertainty impacting the retrieval.In a case study of monitoring an extreme aerosol event,the AGRI AOD is found to be able to capture the detailed temporal evolution of the event.This work demonstrates the superiority of the transfer-learning technique in satellite AOD retrievals and the applicability of the retrieved AGRI AOD in monitoring extreme pollution events.

An Opposition-Based Learning-Based Search Mechanism for Flying Foxes Optimization Algorithm

The flying foxes optimization(FFO)algorithm,as a newly introduced metaheuristic algorithm,is inspired by the survival tactics of flying foxes in heat wave environments.FFO preferentially selects the best-performing individuals.This tendency will cause the newly generated solution to remain closely tied to the candidate optimal in the search area.To address this issue,the paper introduces an opposition-based learning-based search mechanism for FFO algorithm(IFFO).Firstly,this paper introduces niching techniques to improve the survival list method,which not only focuses on the adaptability of individuals but also considers the population’s crowding degree to enhance the global search capability.Secondly,an initialization strategy of opposition-based learning is used to perturb the initial population and elevate its quality.Finally,to verify the superiority of the improved search mechanism,IFFO,FFO and the cutting-edge metaheuristic algorithms are compared and analyzed using a set of test functions.The results prove that compared with other algorithms,IFFO is characterized by its rapid convergence,precise results and robust stability.

An Online Fake Review Detection Approach Using Famous Machine Learning Algorithms

Online review platforms are becoming increasingly popular,encouraging dishonest merchants and service providers to deceive customers by creating fake reviews for their goods or services.Using Sybil accounts,bot farms,and real account purchases,immoral actors demonize rivals and advertise their goods.Most academic and industry efforts have been aimed at detecting fake/fraudulent product or service evaluations for years.The primary hurdle to identifying fraudulent reviews is the lack of a reliable means to distinguish fraudulent reviews from real ones.This paper adopts a semi-supervised machine learning method to detect fake reviews on any website,among other things.Online reviews are classified using a semi-supervised approach(PU-learning)since there is a shortage of labeled data,and they are dynamic.Then,classification is performed using the machine learning techniques Support Vector Machine(SVM)and Nave Bayes.The performance of the suggested system has been compared with standard works,and experimental findings are assessed using several assessment metrics.

Data-Driven Learning Control Algorithms for Unachievable Tracking Problems

For unachievable tracking problems, where the system output cannot precisely track a given reference, achieving the best possible approximation for the reference trajectory becomes the objective. This study aims to investigate solutions using the Ptype learning control scheme. Initially, we demonstrate the necessity of gradient information for achieving the best approximation.Subsequently, we propose an input-output-driven learning gain design to handle the imprecise gradients of a class of uncertain systems. However, it is discovered that the desired performance may not be attainable when faced with incomplete information.To address this issue, an extended iterative learning control scheme is introduced. In this scheme, the tracking errors are modified through output data sampling, which incorporates lowmemory footprints and offers flexibility in learning gain design.The input sequence is shown to converge towards the desired input, resulting in an output that is closest to the given reference in the least square sense. Numerical simulations are provided to validate the theoretical findings.

Unleashing the Power of Multi-Agent Reinforcement Learning for Algorithmic Trading in the Digital Financial Frontier and Enterprise Information Systems

In the rapidly evolving landscape of today’s digital economy,Financial Technology(Fintech)emerges as a trans-formative force,propelled by the dynamic synergy between Artificial Intelligence(AI)and Algorithmic Trading.Our in-depth investigation delves into the intricacies of merging Multi-Agent Reinforcement Learning(MARL)and Explainable AI(XAI)within Fintech,aiming to refine Algorithmic Trading strategies.Through meticulous examination,we uncover the nuanced interactions of AI-driven agents as they collaborate and compete within the financial realm,employing sophisticated deep learning techniques to enhance the clarity and adaptability of trading decisions.These AI-infused Fintech platforms harness collective intelligence to unearth trends,mitigate risks,and provide tailored financial guidance,fostering benefits for individuals and enterprises navigating the digital landscape.Our research holds the potential to revolutionize finance,opening doors to fresh avenues for investment and asset management in the digital age.Additionally,our statistical evaluation yields encouraging results,with metrics such as Accuracy=0.85,Precision=0.88,and F1 Score=0.86,reaffirming the efficacy of our approach within Fintech and emphasizing its reliability and innovative prowess.

Uniaxial Compressive Strength Prediction for Rock Material in Deep Mine Using Boosting-Based Machine Learning Methods and Optimization Algorithms

Traditional laboratory tests for measuring rock uniaxial compressive strength(UCS)are tedious and timeconsuming.There is a pressing need for more effective methods to determine rock UCS,especially in deep mining environments under high in-situ stress.Thus,this study aims to develop an advanced model for predicting the UCS of rockmaterial in deepmining environments by combining three boosting-basedmachine learning methods with four optimization algorithms.For this purpose,the Lead-Zinc mine in Southwest China is considered as the case study.Rock density,P-wave velocity,and point load strength index are used as input variables,and UCS is regarded as the output.Subsequently,twelve hybrid predictive models are obtained.Root mean square error(RMSE),mean absolute error(MAE),coefficient of determination(R2),and the proportion of the mean absolute percentage error less than 20%(A-20)are selected as the evaluation metrics.Experimental results showed that the hybridmodel consisting of the extreme gradient boostingmethod and the artificial bee colony algorithm(XGBoost-ABC)achieved satisfactory results on the training dataset and exhibited the best generalization performance on the testing dataset.The values of R2,A-20,RMSE,and MAE on the training dataset are 0.98,1.0,3.11 MPa,and 2.23MPa,respectively.The highest values of R2 and A-20(0.93 and 0.96),and the smallest RMSE and MAE values of 4.78 MPa and 3.76MPa,are observed on the testing dataset.The proposed hybrid model can be considered a reliable and effective method for predicting rock UCS in deep mines.

Marine Predators Algorithm with Deep Learning-Based Leukemia Cancer Classification on Medical Images

In blood or bone marrow,leukemia is a form of cancer.A person with leukemia has an expansion of white blood cells(WBCs).It primarily affects children and rarely affects adults.Treatment depends on the type of leukemia and the extent to which cancer has established throughout the body.Identifying leukemia in the initial stage is vital to providing timely patient care.Medical image-analysis-related approaches grant safer,quicker,and less costly solutions while ignoring the difficulties of these invasive processes.It can be simple to generalize Computer vision(CV)-based and image-processing techniques and eradicate human error.Many researchers have implemented computer-aided diagnosticmethods andmachine learning(ML)for laboratory image analysis,hopefully overcoming the limitations of late leukemia detection and determining its subgroups.This study establishes a Marine Predators Algorithm with Deep Learning Leukemia Cancer Classification(MPADL-LCC)algorithm onMedical Images.The projectedMPADL-LCC system uses a bilateral filtering(BF)technique to pre-process medical images.The MPADL-LCC system uses Faster SqueezeNet withMarine Predators Algorithm(MPA)as a hyperparameter optimizer for feature extraction.Lastly,the denoising autoencoder(DAE)methodology can be executed to accurately detect and classify leukemia cancer.The hyperparameter tuning process using MPA helps enhance leukemia cancer classification performance.Simulation results are compared with other recent approaches concerning various measurements and the MPADL-LCC algorithm exhibits the best results over other recent approaches.

Quantification of the concrete freeze–thaw environment across the Qinghai–Tibet Plateau based on machine learning algorithms

The reasonable quantification of the concrete freezing environment on the Qinghai–Tibet Plateau(QTP) is the primary issue in frost resistant concrete design, which is one of the challenges that the QTP engineering managers should take into account. In this paper, we propose a more realistic method to calculate the number of concrete freeze–thaw cycles(NFTCs) on the QTP. The calculated results show that the NFTCs increase as the altitude of the meteorological station increases with the average NFTCs being 208.7. Four machine learning methods, i.e., the random forest(RF) model, generalized boosting method(GBM), generalized linear model(GLM), and generalized additive model(GAM), are used to fit the NFTCs. The root mean square error(RMSE) values of the RF, GBM, GLM, and GAM are 32.3, 4.3, 247.9, and 161.3, respectively. The R^(2) values of the RF, GBM, GLM, and GAM are 0.93, 0.99, 0.48, and 0.66, respectively. The GBM method performs the best compared to the other three methods, which was shown by the results of RMSE and R^(2) values. The quantitative results from the GBM method indicate that the lowest, medium, and highest NFTC values are distributed in the northern, central, and southern parts of the QTP, respectively. The annual NFTCs in the QTP region are mainly concentrated at 160 and above, and the average NFTCs is 200 across the QTP. Our results can provide scientific guidance and a theoretical basis for the freezing resistance design of concrete in various projects on the QTP.

Gradient Optimizer Algorithm with Hybrid Deep Learning Based Failure Detection and Classification in the Industrial Environment

Failure detection is an essential task in industrial systems for preventing costly downtime and ensuring the seamlessoperation of the system. Current industrial processes are getting smarter with the emergence of Industry 4.0.Specifically, various modernized industrial processes have been equipped with quite a few sensors to collectprocess-based data to find faults arising or prevailing in processes along with monitoring the status of processes.Fault diagnosis of rotating machines serves a main role in the engineering field and industrial production. Dueto the disadvantages of existing fault, diagnosis approaches, which greatly depend on professional experienceand human knowledge, intellectual fault diagnosis based on deep learning (DL) has attracted the researcher’sinterest. DL reaches the desired fault classification and automatic feature learning. Therefore, this article designs a Gradient Optimizer Algorithm with Hybrid Deep Learning-based Failure Detection and Classification (GOAHDLFDC)in the industrial environment. The presented GOAHDL-FDC technique initially applies continuous wavelettransform (CWT) for preprocessing the actual vibrational signals of the rotating machinery. Next, the residualnetwork (ResNet18) model was exploited for the extraction of features from the vibration signals which are thenfed into theHDLmodel for automated fault detection. Finally, theGOA-based hyperparameter tuning is performedtoadjust the parameter valuesof theHDLmodel accurately.The experimental result analysis of the GOAHDL-FD Calgorithm takes place using a series of simulations and the experimentation outcomes highlight the better resultsof the GOAHDL-FDC technique under different aspects.

Extended Deep Learning Algorithm for Improved Brain Tumor Diagnosis System

At present,the prediction of brain tumors is performed using Machine Learning(ML)and Deep Learning(DL)algorithms.Although various ML and DL algorithms are adapted to predict brain tumors to some range,some concerns still need enhancement,particularly accuracy,sensitivity,false positive and false negative,to improve the brain tumor prediction system symmetrically.Therefore,this work proposed an Extended Deep Learning Algorithm(EDLA)to measure performance parameters such as accuracy,sensitivity,and false positive and false negative rates.In addition,these iterated measures were analyzed by comparing the EDLA method with the Convolutional Neural Network(CNN)way further using the SPSS tool,and respective graphical illustrations were shown.The results were that the mean performance measures for the proposed EDLA algorithm were calculated,and those measured were accuracy(97.665%),sensitivity(97.939%),false positive(3.012%),and false negative(3.182%)for ten iterations.Whereas in the case of the CNN,the algorithm means accuracy gained was 94.287%,mean sensitivity 95.612%,mean false positive 5.328%,and mean false negative 4.756%.These results show that the proposed EDLA method has outperformed existing algorithms,including CNN,and ensures symmetrically improved parameters.Thus EDLA algorithm introduces novelty concerning its performance and particular activation function.This proposed method will be utilized effectively in brain tumor detection in a precise and accurate manner.This algorithm would apply to brain tumor diagnosis and be involved in various medical diagnoses aftermodification.If the quantity of dataset records is enormous,then themethod’s computation power has to be updated.

改进Q-Learning的路径规划算法研究

针对Q-Learning算法学习效率低、收敛速度慢且在动态障碍物的环境下路径规划效果不佳的问题,本文提出一种改进Q-Learning的移动机器人路径规划算法.针对该问题,算法根据概率的突变性引入探索因子来平衡探索和利用以加快学习效率;通过在更新函数中设计深度学习因子以保证算法探索概率;融合遗传算法,避免陷入局部路径最优同时按阶段探索最优迭代步长次数,以减少动态地图探索重复率;最后提取输出的最优路径关键节点采用贝塞尔曲线进行平滑处理,进一步保证路径平滑度和可行性.实验通过栅格法构建地图,对比实验结果表明,改进后的算法效率相较于传统算法在迭代次数和路径上均有较大优化,且能够较好的实现动态地图下的路径规划,进一步验证所提方法的有效性和实用性.

基于改进Q-Learning的移动机器人路径规划算法

随着移动机器人在生产生活中的深入应用,其路径规划能力也需要向快速性和环境适应性兼备发展。为解决现有移动机器人使用强化学习方法进行路径规划时存在的探索前期容易陷入局部最优、反复搜索同一区域,探索后期收敛率低、收敛速度慢的问题,本研究提出一种改进的Q-Learning算法。该算法改进Q矩阵赋值方法,使迭代前期探索过程具有指向性,并降低碰撞的情况;改进Q矩阵迭代方法,使Q矩阵更新具有前瞻性,避免在一个小区域中反复探索;改进随机探索策略,在迭代前期全面利用环境信息,后期向目标点靠近。在不同栅格地图仿真验证结果表明,本文算法在Q-Learning算法的基础上,通过上述改进降低探索过程中的路径长度、减少抖动并提高收敛的速度,具有更高的计算效率。

Advancing automated pupillometry:a practical deep learning model utilizing infrared pupil images

AIM:To establish pupil diameter measurement algorithms based on infrared images that can be used in real-world clinical settings.METHODS:A total of 188 patients from outpatient clinic at He Eye Specialist Shenyang Hospital from Spetember to December 2022 were included,and 13470 infrared pupil images were collected for the study.All infrared images for pupil segmentation were labeled using the Labelme software.The computation of pupil diameter is divided into four steps:image pre-processing,pupil identification and localization,pupil segmentation,and diameter calculation.Two major models are used in the computation process:the modified YoloV3 and Deeplabv 3+models,which must be trained beforehand.RESULTS:The test dataset included 1348 infrared pupil images.On the test dataset,the modified YoloV3 model had a detection rate of 99.98% and an average precision(AP)of 0.80 for pupils.The DeeplabV3+model achieved a background intersection over union(IOU)of 99.23%,a pupil IOU of 93.81%,and a mean IOU of 96.52%.The pupil diameters in the test dataset ranged from 20 to 56 pixels,with a mean of 36.06±6.85 pixels.The absolute error in pupil diameters between predicted and actual values ranged from 0 to 7 pixels,with a mean absolute error(MAE)of 1.06±0.96 pixels.CONCLUSION:This study successfully demonstrates a robust infrared image-based pupil diameter measurement algorithm,proven to be highly accurate and reliable for clinical application.

Intelligent Design of High Strength and High Conductivity Copper Alloys Using Machine Learning Assisted by Genetic Algor

Metallic alloys for a given application are usually designed to achieve the desired properties by devising experimentsbased on experience, thermodynamic and kinetic principles, and various modeling and simulation exercises.However, the influence of process parameters and material properties is often non-linear and non-colligative. Inrecent years, machine learning (ML) has emerged as a promising tool to dealwith the complex interrelation betweencomposition, properties, and process parameters to facilitate accelerated discovery and development of new alloysand functionalities. In this study, we adopt an ML-based approach, coupled with genetic algorithm (GA) principles,to design novel copper alloys for achieving seemingly contradictory targets of high strength and high electricalconductivity. Initially, we establish a correlation between the alloy composition (binary to multi-component) andthe target properties, namely, electrical conductivity and mechanical strength. Catboost, an ML model coupledwith GA, was used for this task. The accuracy of the model was above 93.5%. Next, for obtaining the optimizedcompositions the outputs fromthe initial model were refined by combining the concepts of data augmentation andPareto front. Finally, the ultimate objective of predicting the target composition that would deliver the desired rangeof properties was achieved by developing an advancedMLmodel through data segregation and data augmentation.To examine the reliability of this model, results were rigorously compared and verified using several independentdata reported in the literature. This comparison substantiates that the results predicted by our model regarding thevariation of conductivity and evolution ofmicrostructure and mechanical properties with composition are in goodagreement with the reports published in the literature.

Dynamic plugging regulating strategy of pipeline robot based on reinforcement learning

Pipeline isolation plugging robot (PIPR) is an important tool in pipeline maintenance operation. During the plugging process, the violent vibration will occur by the flow field, which can cause serious damage to the pipeline and PIPR. In this paper, we propose a dynamic regulating strategy to reduce the plugging-induced vibration by regulating the spoiler angle and plugging velocity. Firstly, the dynamic plugging simulation and experiment are performed to study the flow field changes during dynamic plugging. And the pressure difference is proposed to evaluate the degree of flow field vibration. Secondly, the mathematical models of pressure difference with plugging states and spoiler angles are established based on the extreme learning machine (ELM) optimized by improved sparrow search algorithm (ISSA). Finally, a modified Q-learning algorithm based on simulated annealing is applied to determine the optimal strategy for the spoiler angle and plugging velocity in real time. The results show that the proposed method can reduce the plugging-induced vibration by 19.9% and 32.7% on average, compared with single-regulating methods. This study can effectively ensure the stability of the plugging process.

AWeb Application Fingerprint Recognition Method Based on Machine Learning

Web application fingerprint recognition is an effective security technology designed to identify and classify web applications,thereby enhancing the detection of potential threats and attacks.Traditional fingerprint recognition methods,which rely on preannotated feature matching,face inherent limitations due to the ever-evolving nature and diverse landscape of web applications.In response to these challenges,this work proposes an innovative web application fingerprint recognition method founded on clustering techniques.The method involves extensive data collection from the Tranco List,employing adjusted feature selection built upon Wappalyzer and noise reduction through truncated SVD dimensionality reduction.The core of the methodology lies in the application of the unsupervised OPTICS clustering algorithm,eliminating the need for preannotated labels.By transforming web applications into feature vectors and leveraging clustering algorithms,our approach accurately categorizes diverse web applications,providing comprehensive and precise fingerprint recognition.The experimental results,which are obtained on a dataset featuring various web application types,affirm the efficacy of the method,demonstrating its ability to achieve high accuracy and broad coverage.This novel approach not only distinguishes between different web application types effectively but also demonstrates superiority in terms of classification accuracy and coverage,offering a robust solution to the challenges of web application fingerprint recognition.

A Two-Layer Encoding Learning Swarm Optimizer Based on Frequent Itemsets for Sparse Large-Scale Multi-Objective Optimization

Traditional large-scale multi-objective optimization algorithms(LSMOEAs)encounter difficulties when dealing with sparse large-scale multi-objective optimization problems(SLM-OPs)where most decision variables are zero.As a result,many algorithms use a two-layer encoding approach to optimize binary variable Mask and real variable Dec separately.Nevertheless,existing optimizers often focus on locating non-zero variable posi-tions to optimize the binary variables Mask.However,approxi-mating the sparse distribution of real Pareto optimal solutions does not necessarily mean that the objective function is optimized.In data mining,it is common to mine frequent itemsets appear-ing together in a dataset to reveal the correlation between data.Inspired by this,we propose a novel two-layer encoding learning swarm optimizer based on frequent itemsets(TELSO)to address these SLMOPs.TELSO mined the frequent terms of multiple particles with better target values to find mask combinations that can obtain better objective values for fast convergence.Experi-mental results on five real-world problems and eight benchmark sets demonstrate that TELSO outperforms existing state-of-the-art sparse large-scale multi-objective evolutionary algorithms(SLMOEAs)in terms of performance and convergence speed.

Constrained Multi-Objective Optimization With Deep Reinforcement Learning Assisted Operator Selection

Solving constrained multi-objective optimization problems with evolutionary algorithms has attracted considerable attention.Various constrained multi-objective optimization evolutionary algorithms(CMOEAs)have been developed with the use of different algorithmic strategies,evolutionary operators,and constraint-handling techniques.The performance of CMOEAs may be heavily dependent on the operators used,however,it is usually difficult to select suitable operators for the problem at hand.Hence,improving operator selection is promising and necessary for CMOEAs.This work proposes an online operator selection framework assisted by Deep Reinforcement Learning.The dynamics of the population,including convergence,diversity,and feasibility,are regarded as the state;the candidate operators are considered as actions;and the improvement of the population state is treated as the reward.By using a Q-network to learn a policy to estimate the Q-values of all actions,the proposed approach can adaptively select an operator that maximizes the improvement of the population according to the current state and thereby improve the algorithmic performance.The framework is embedded into four popular CMOEAs and assessed on 42 benchmark problems.The experimental results reveal that the proposed Deep Reinforcement Learning-assisted operator selection significantly improves the performance of these CMOEAs and the resulting algorithm obtains better versatility compared to nine state-of-the-art CMOEAs.

A Whale Optimization Algorithm with Distributed Collaboration and Reverse Learning Ability

Due to the development of digital transformation,intelligent algorithms are getting more and more attention.The whale optimization algorithm(WOA)is one of swarm intelligence optimization algorithms and is widely used to solve practical engineering optimization problems.However,with the increased dimensions,higher requirements are put forward for algorithm performance.The double population whale optimization algorithm with distributed collaboration and reverse learning ability(DCRWOA)is proposed to solve the slow convergence speed and unstable search accuracy of the WOA algorithm in optimization problems.In the DCRWOA algorithm,the novel double population search strategy is constructed.Meanwhile,the reverse learning strategy is adopted in the population search process to help individuals quickly jump out of the non-ideal search area.Numerical experi-ments are carried out using standard test functions with different dimensions(10,50,100,200).The optimization case of shield construction parameters is also used to test the practical application performance of the proposed algo-rithm.The results show that the DCRWOA algorithm has higher optimization accuracy and stability,and the convergence speed is significantly improved.Therefore,the proposed DCRWOA algorithm provides a better method for solving practical optimization problems.