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.

Do Search and Selection Operators Play Important Roles in Multi-Objective Evolutionary Algorithms:A Case Study

Multi-objective Evolutionary Algorithm (MOEA) is becoming a hot research area and quite a few aspects of MOEAs have been studied and discussed. However there are still few literatures discussing the roles of search and selection operators in MOEAs. This paper studied their roles by solving a case of discrete Multi-objective Optimization Problem (MOP): Multi-objective TSP with a new MOEA. In the new MOEA, We adopt an efficient search operator, which has the properties of both crossover and mutation, to generate the new individuals and chose two selection operators: Family Competition and Population Competition with probabilities to realize selection. The simulation experiments showed that this new MOEA could get good uniform solutions representing the Pareto Front and outperformed SPEA in almost every simulation run on this problem. Furthermore, we analyzed its convergence property using finite Markov chain and proved that it could converge to Pareto Front with probability 1. We also find that the convergence property of MOEAs has much relationship with search and selection operators.

Characterization of immunogenic cell death-related genes predicting prognosis in colon adenocarcinoma

Background:Colon adenocarcinoma(COAD)is a gastrointestinal malignancy with a high mortality rate.Studies have confirmed the role of immunogenic cell death(ICD)in different cancer types.However,there is a lack of research on ICD-related genes(ICD-RGs)in COAD.This study aimed to examine the impact of ICD-RGs on COAD and their interaction with the immune microenvironment.Methods:Using data from The Cancer Genome Atlas and Gene Expression Omnibus databases,we identified 107 ICD-RGs in COAD.Using a one-way Cox regression analysis,we examined the relationship between these ICD-RGs and overall survival in COAD.Results:Following the regression analyses,we identified 14 overall survival-related genes.Furthermore,we examined the predictive impact of the ICD-RGs using the least absolute shrinkage and selection operator regression analysis and developed a nine-genes prognostic model.The Cancer Genome Atlas and Gene Expression Omnibus datasets were used for training and validation.Kaplan-Meier analysis was used to confirm that the high-risk group had a lower survival rate than the low-risk group.Finally,following a multifactorial analysis,we created a prognostic nomogram that integrated clinical data and risk scores.Conclusions:The nine-genes model exhibits robust stability and can provide valuable insights for guiding the development of tumor immunotherapy strategies and personalized drug selection for patients with COAD.

Learning to select the recombination operator for derivative-free optimization

Extensive studies on selecting recombination operators adaptively,namely,adaptive operator selection(AOS),during the search process of an evolutionary algorithm(EA),have shown that AOS is promising for improving EA's performance.A variety of heuristic mechanisms for AOS have been proposed in recent decades,which usually contain two main components:the feature extraction and the policy setting.The feature extraction refers to as extracting relevant features from the information collected during the search process.The policy setting means to set a strategy(or policy)on how to select an operator from a pool of operators based on the extracted feature.Both components are designed by hand in existing studies,which may not be efficient for adapting optimization problems.In this paper,a generalized framework is proposed for learning the components of AOS for one of the main streams of EAs,namely,differential evolution(DE).In the framework,the feature extraction is parameterized as a deep neural network(DNN),while a Dirichlet distribution is considered to be the policy.A reinforcement learning method,named policy gradient,is used to train the DNN.As case studies,the proposed framework is applied to two DEs including the classic DE and a recently-proposed DE,which result in two new algorithms named PG-DE and PG-MPEDE,respectively.Experiments on the Congress of Evolutionary Computation(CEC)2018 test suite show that the proposed new algorithms perform significantly better than their counterparts.Finally,we prove theoretically that the considered classic methods are the special cases of the proposed framework.

Identification of an immune classifier for predicting the prognosis and therapeutic response in triple-negative breast cancer

Triple-negative breast cancer(TNBC)poses a significant challenge due to the lack of reliable prognostic gene signatures and an understanding of its immune behavior.Methods:We analyzed clinical information and mRNA expression data from 162 TNBC patients in TCGA-BRCA and 320 patients in METABRIC-BRCA.Utilizing weighted gene coexpression network analysis,we pinpointed 34 TNBC immune genes linked to survival.The least absolute shrinkage and selection operator Cox regression method identified key TNBC immune candidates for prognosis prediction.We calculated chemotherapy sensitivity scores using the“pRRophetic”package in R software and assessed immunotherapy response using the Tumor Immune Dysfunction and Exclusion algorithm.Results:In this study,34 survival-related TNBC immune gene expression profiles were identified.A least absolute shrinkage and selection operator-Cox regression model was used and 15 candidates were prioritized,with a concomitant establishment of a robust risk immune classifier.The high-risk TNBC immune groups showed increased sensitivity to therapeutic agents like RO-3306,Tamoxifen,Sunitinib,JNK Inhibitor VIII,XMD11-85h,BX-912,and Tivozanib.An analysis of the Search Tool for Interaction of Chemicals database revealed the associations between the high-risk group and signaling pathways,such as those involving Rap1,Ras,and PI3K-Akt.The low-risk group showed a higher immunotherapy response rate,as observed through the tumor immune dysfunction and exclusion analysis in the TCGA-TNBC and METABRIC-TNBC cohorts.Conclusion:This study provides insights into the immune complexities of TNBC,paving the way for novel diagnostic approaches and precision treatment methods that exploit its immunological intricacies,thus offering hope for improved management and outcomes of this challenging disease.

Integrated analysis of single-cell and bulk RNA-seq establishes a novel signature for prediction in gastric cancer

BACKGROUND Single-cell sequencing technology provides the capability to analyze changes in specific cell types during the progression of disease.However,previous single-cell sequencing studies on gastric cancer(GC)have largely focused on immune cells and stromal cells,and further elucidation is required regarding the alterations that occur in gastric epithelial cells during the development of GC.AIM To create a GC prediction model based on single-cell and bulk RNA sequencing(bulk RNA-seq)data.METHODS In this study,we conducted a comprehensive analysis by integrating three singlecell RNA sequencing(scRNA-seq)datasets and ten bulk RNA-seq datasets.Our analysis mainly focused on determining cell proportions and identifying differentially expressed genes(DEGs).Specifically,we performed differential expression analysis among epithelial cells in GC tissues and normal gastric tissues(NAGs)and utilized both single-cell and bulk RNA-seq data to establish a prediction model for GC.We further validated the accuracy of the GC prediction model in bulk RNA-seq data.We also used Kaplan–Meier plots to verify the correlation between genes in the prediction model and the prognosis of GC.RESULTS By analyzing scRNA-seq data from a total of 70707 cells from GC tissue,NAG,and chronic gastric tissue,10 cell types were identified,and DEGs in GC and normal epithelial cells were screened.After determining the DEGs in GC and normal gastric samples identified by bulk RNA-seq data,a GC predictive classifier was constructed using the Least absolute shrinkage and selection operator(LASSO)and random forest methods.The LASSO classifier showed good performance in both validation and model verification using The Cancer Genome Atlas and Genotype-Tissue Expression(GTEx)datasets[area under the curve(AUC)_min=0.988,AUC_1se=0.994],and the random forest model also achieved good results with the validation set(AUC=0.92).Genes TIMP1,PLOD3,CKS2,TYMP,TNFRSF10B,CPNE1,GDF15,BCAP31,and CLDN7 were identified to have high importance values in multiple GC predictive models,and KM-PLOTTER analysis showed their relevance to GC prognosis,suggesting their potential for use in GC diagnosis and treatment.CONCLUSION A predictive classifier was established based on the analysis of RNA-seq data,and the genes in it are expected to serve as auxiliary markers in the clinical diagnosis of GC.

Optimization of maintenance strategy for high-speed railwaycatenary system based on multistate model

A multi-objective optimization model considering both reliability and maintenance cost is proposed to solve the contradiction between reliability and maintenance cost in high-speed railway catenary system maintenance activities.The non-dominated sorting genetic algorithm 2(NSGA2)is applied to multi-objective optimization,and the optimization result is a set of Pareto solutions.Firstly,multistate failure mode analysis is conducted for the main devices leading to the failure of catenary,and then the reliability and failure mode of the whole catenary system is analyzed.The mathematical relationship between system reliability and maintenance cost is derived considering the existing catenary preventive maintenance mode to improve the reliability of the system.Secondly,an improved NSGA2(INSGA2)is proposed,which strengths population diversity by improving selection operator,and introduces local search strategy to ensure that population distribution is more uniform.The comparison results of the two algorithms before and after improvement on the zero-ductility transition(ZDT)series functions show that the population diversity is better and the solution is more uniform using INSGA2.Finally,the INSGA2 is applied to multi-objective optimization of system reliability and maintenance cost in different maintenance periods.The decision-makers can choose the reasonable solutions as the maintenance plans in the optimization results by weighing the relationship between the system reliability and the maintenance cost.The selected maintenance plans can ensure the lowest maintenance cost while the system reliability is as high as possible.

Discrimination of Acori Tatarinowii Rhizoma from two habitats based on GC-MS fingerprinting and LASSO-PLS-DA

This study is intended to explore the chemical differences of Acori Tatarinowii Rhizoma （ATR） samples collected from two habitats, Sichuan and Anhui provinces, China. Gas chromatography-mass spectrometry （GC-MS） was applied to establishing the quantitative chemical fingerprints of ATRs. A total of 104 volatile compounds were identified and quantified with the information of mass spectra and retention index （RI）. Furthermore, least absolute shrinkage and selection operator （LASSO）, a sparse regularization method, combined with subsampling was employed to improve the classification ability of partial least squares-discriminant analysis （PLS-DA）. After variable selection by LASSO, three chemical markers,β-elemene, α-selinene and α-asarone, were identified for the discrimination of ATRs from two habitats, and the total classification correct rate was increased from 82.76% to 96.55%. The proposed LASSO-PLS-DA method can serve as an efficient strategy for screening marked chemical components and geo-herbalism research of traditional Chinese medicines.

Sparse reconstruction for fluorescence molecular tomography via a fast iterative algorithm

Fluorescence molecular tomography(FMT)is a fast-developing optical imaging modalitythat has great potential in early diagnosis of disease and drugs development.However,recon-struction algorithms have to address a highly ill-posed problem to fulfll 3D reconstruction inFMT.In this contribution,we propose an efficient iterative algorithm to solve the large-scalereconstruction problem,in which the sparsity of fluorescent targets is taken as useful a prioriinformation in designing the reconstruction algorithm.In the implementation,a fast sparseapproximation scheme combined with a stage-wise learning strategy enable the algorithm to dealwith the ill-posed inverse problem at reduced computational costs.We validate the proposed fastiterative method with numerical simulation on a digital mouse model.Experimental results demonstrate that our method is robust for different finite element meshes and different Poissonnoise levels.

Study on Influence Factors of Pressure Injury Risk in the Elderly Inpatients with Kidney Disease Based on LASSO Regression

Objective: This paper aims to explore clinical status and related influence factors of pressure injury (PI) in the elderly inpatients with kidney disease, so as to provide reference for the prevention and treatment of PI in the elderly inpatients with kidney disease. Methods: Retrospective collection method is adopted to collect 158 clinical cases of the elderly inpatients with kidney disease aged ≥ 60 in the Nephrology Department, the First Affiliated Hospital of Jinan University from January 2017 to December 2019, and then least absolute shrinkage and selection Operator (LASSO) regression analysis is used to analyze 17 possible influence factors;finally Logistic regression model is established to analyze and screen influence factors of risk. Results: 1) Among 158 elderly inpatients with medium and high risk of PI, the incidence of PI is 20.25%;the most common stage of injury is stage I (42.5%);sacrococcygeal (60%) is the high-risk site of pressure injury. 2) LASSO regression analysis shows that history of present respiratory infection/respiratory failure (β = 1.2714. P < 0.05) and hospitalization time (β = 0.4177. P < 0.05) are independent factors influencing PI risk in the elderly inpatients with kidney disease. Conclusion: The elderly patients with kidney disease and PI risk are the high incidence population of hospital acquired PI;for the elderly inpatients with kidney disease and having respiratory infection history or respiratory failure, prolonged hospitalization will significantly increase the risk of PI. Therefore, targeted preventive and control measures should be taken to reduce the incidence of PI.

Development and validation of a three-long noncoding RNA signature for predicting prognosis of patients with gastric cancer

BACKGROUND Gastric cancer(GC)is one of the most frequently diagnosed gastrointestinal cancers throughout the world.Novel prognostic biomarkers are required to predict the prognosis of GC.AIM To identify a multi-long noncoding RNA(lncRNA)prognostic model for GC.METHODS Transcriptome data and clinical data were downloaded from The Cancer Genome Atlas.COX and least absolute shrinkage and selection operator regression analyses were performed to screen for prognosis associated lncRNAs.Receiver operating characteristic curve and Kaplan-Meier survival analyses were applied to evaluate the effectiveness of the model.RESULTS The prediction model was established based on the expression of AC007991.4,AC079385.3,and AL109615.2 Based on the model,GC patients were divided into“high risk”and“low risk”groups to compare the differences in survival.The model was re-evaluated with the clinical data of our center.CONCLUSION The 3-lncRNA combination model is an independent prognostic factor for GC.

The Properties Analysis for Generalized Abstract Evolutionary Algorithm

There has been a growing interest in mathematical models to character the evolutionary algorithms. The best-known one of such models is the axiomatic model called the abstract evolutionary algorithm （AEA）, which unifies most of the currently known evolutionary algorithms and describes the evolution as an abstract stochastic process composed of two fundamental abstract operators： abstract selection and evolution operators. In this paper, we first introduce the definitions of the generalized abstract selection and evolution operators. Then we discuss the characterization of some parameters related to generalized abstract selection and evolution operators. Based on these operators, we finally give the strong convergence of the generalized abstract evolutionary algorithm. The present work provides a big step toward the establishment of a unified theory of evolutionary computation.

A NOTE ON PROBABILISTIC CONTRACTOR COUPLE AND SOLUTIONS FOR A SYSTEM OF NONLINEAR EQUATIONS IN N A MENGER PN-SPACES

The concept of (Phi, Delta)-type probabilistic contractor couple was introduced which simplifies and weakens the definition of probabilistic contractor couple given by Zhang Shisheng. The existence and uniqueness of the solutions for a system of nonlinear operator equations with this kind of propabilistic contractor couple in N. A. Menger PN-spaces were studied. The works improve and extend the corresponding results by M. Altman, A. C. Lee, W. J. Padgett et al.

Based on necroptosis identifying the immunological features and prognostic signatures of lung adenocarcinoma

Background:Lung adenocarcinoma is one of the most common pathological types of lung malignant tumor with high morbidity and mortality.Long non-coding RNAs are gradually recognized to play crucial roles in tumor occurrence and development.Necroptosis is a newly established way for cell programmed death,undertaking essential roles in anti-tumor.Therefore,identifying necroptosis-related l ong non-coding RNAs and based on them to evaluate the signatures of l ung adenocarcinoma is essential for patients’survival prediction and therapy.Methods:We collected data from the public database and performed the least absolute shrinkage to construct a 13-lncRNAs prognostic model.Based on the Consensus Clustering,ESTIMATE,CIRERSORT,and weighted gene co-expression network analysis to identify the immune signatures.Results:This study identified a 13-lncRNAs prognostic model.The model’s prediction accuracy was evaluated by receiver operating characteristic and independent-prognosis analysis;besides,a Gene Expression Omnibus dataset was applied for external validation.Furthermore,we analyzed the immune features of subgroups in multiple dimensions.A consensus clustering analysis based on the 41 genes was implemented to separate lung adenocarcinoma patients into two subgroups.We compared the features of subgroups in multiple dimensions,including survival,immune microenvironment,immune cells infiltration and gene co-expression network analysis.Conclusion:W e established a prognosis necroptosis-related risk model to predict lung adenocarcinoma patients’prognosis and systematically understood the correlation between immune and necroptosis.This study can applicate in clinical to predict the prognosis of lung adenocarcinoma patients and provide new insight into lung adenocarcinoma immune therapy.

Circulating biomarker-and magnetic resonance-based nomogram predicting long-term outcomes in dilated cardiomyopathy

Background:Dilated cardiomyopathy(DCM)has a high mortality rate and is the most common indication for heart transplantation.Our study sought to develop a multiparametric nomogram to assess individualized all-cause mortality or heart transplantation(ACM/HTx)risk in DCM patients.Methods:The present study is a retrospective cohort study.The demographic,clinical,blood test,and cardiac magnetic resonance imaging(CMRI)data of DCM patients in the tertiary center(Fuwai Hospital)were collected.The primary endpoint was ACM/HTx.The least absolute shrinkage and selection operator(LASSO)Cox regression model was applied for variable selection.Multivariable Cox regression was used to develop a nomogram.The concordance index(C-index),area under the receiver operating characteristic curve(AUC),calibration curve,and decision curve analysis(DCA)were used to evaluate the performance of the nomogram.Results:A total of 218 patients were included in the present study.They were randomly divided into a training cohort and a validation cohort.The nomogram was established based on eight variables,including mid-wall late gadolinium enhancement,systolic blood pressure,diastolic blood pressure,left ventricular ejection fraction,left ventricular end-diastolic diameter,left ventricular end-diastolic volume index,free triiodothyronine,and N-terminal pro-B type natriuretic peptide.The AUCs regarding 1-year,3-year,and 5-year ACM/HTx events were 0.859,0.831,and 0.840 in the training cohort and 0.770,0.789,and 0.819 in the validation cohort,respectively.The calibration curve and DCA showed good accuracy and clinical utility of the nomogram.Conclusions:We established and validated a circulating biomarker-and CMRI-based nomogram that could provide a personalized prediction of ACM/HTx for DCM patients,which might help risk stratification and decision-making in clinical practice.

A robust vector tracking loop structure based on potential bias analysis

This paper proposes a robust vector tracking loop structure based on potential bias analysis. The influence of four kinds of biases on the existing two implementations of Vector Tracking Loops(VTLs) is illustrated by theoretical analysis and numerical simulations, and the following findings are obtained. Firstly, the initial user state bias leads to steady navigation solution bias in the relative VTL, while new measurements can eliminate it in the absolute VTL. Secondly, the initial code phase bias is transferred to the following navigation solutions in the relative VTL, while new measurements can eliminate it in the absolute VTL. Thirdly, the user state bias induced by erroneous navigation solution of VTLs can be eliminated by both of the two VTLs. Fourthly,the multipath/NLOS likely affects the two VTLs, and the induced tracking bias in the duration of the multipath/NLOS would decrease the performance of VTLs. Based on the above analysis,a robust VTL structure is proposed, where the absolute VTL is selected for its robustness to the two kinds of initialization biases;meanwhile, the instant bias detection and correction method is used to improve the performance of VTLs in the duration of the multipath/NLOS. Numerical simulations and experimental results verify the effectiveness of the proposed robust VTL structure.

Very Short-Term Probabilistic Prediction Method for Wind Speed Based on ALASSO-Nonlinear Quantile Regression and Integrated Criterion

To enhance the performance of the prediction intervals (PIs), a novel very short-term probabilistic prediction method for wind speed via nonlinear quantile regression (NQR) based on adaptive least absolute shrinkage and selection operator (ALASSO) and integrated criterion (IC) is proposed. The ALASSO method is studied for shrinkage of output weights and selection of variables. Furthermore, for the better performance of PIs, composite weighted linear programming (CWLP) is proposed to modify the conventional linear programming cost function of quantile regression (QR), by combining it with Bayesian information criterion (BIC) as an IC to optimize the coefficients of PIs. Then, the multiple fold cross model (MFCM) is utilized to improve the PIs performance. Multistep probabilistic prediction of 15-minute wind speed is performed based on the real wind farm data from the northeast of China. The effectiveness of the proposed approach is validated through the performances' comparisons with conventional methods.

Aerobic granules cultivated and operated in continuous-flow bioreactor under particle-size selective pressure

A novel method based on the selective pressure of particle size （particle-size cultivation method, PSCM） was developed for the cultivation and operation of aerobic granular sludge in a continuous-flow reactor, and compared with the conventional method based on the selective pressure of settling velocity （settling-velocity cultivation method, SVCM）. Results indicated that aerobic granules could be cultivated in continuous operation mode by this developed method within 14 days. Although in the granulation process, under particle-size selective pressure, mixed liquor suspended solids （MLSS） in the reactor fluctuated greatly and filamentous bacteria dominated the sludge system during the initial operation days, no obvious difference in profile was found between the aerobic granules cultivated by PSCM and SVCM. Moreover, aerobic granules cultivated by PSCM presented larger diameter, lower water content and higher specific rates of nitrification, denitrifieation and phosphorus removal, but lower settling velocity. Under long term operation of more than 30 days, aerobic granules in the continuous-flow reactor could remain stable and obtain good chemical oxygen demand （COD）, NH4^＋-N, total nitrogen （TN） and total phosphorus （TP） removal. The results indicate that PSCM was dependent on the cultivation and maintenance of the stability of aerobic granules in continuous-flow bioreactors.

基于Bayesian Bootstrap抽样的高维线性回归模型

研究小样本下高维线性回归模型中的变量选择问题和模型预测能力。当自变量维数p远大于样本量n时,提出基于Bayesian bootstrap抽样的SCAD(smoothly clipped absolute deviation)压缩方法。仿真和实证分析表明,与SCAD和LASSO(least absolute shrinkage and selection operator)两种传统回归压缩方法相比,本算法受随机干扰影响较小。当样本量较小时,本算法的变量压缩结果更好,变量选择能力更强,模型的标准均方误差值也最小,且模型预测能力提升明显。

Spatial-temporal heterogeneity and determinants of HIV prevalence in the Mano River Union countries

Background:Utilizing population-based survey data in epidemiological research with a spatial perspective can integrate valuable context into the dynamics of HIV prevalence in West Africa.However,the situation in the Mano River Union(MRU)countries is largely unknown.This research aims to perform an ecological study to determine the HIV prevalence patterns in MRU.Methods:We analyzed Demographic and Health Survey(DHS)and AIDS Indicator Survey(AIS)data on HIV prevalence in MRU from 2005 to 2020.We examined the country-specifc,regional-specifc and sex-specifc ratios of respondents to profle the spatial–temporal heterogeneity of HIV prevalence and determine HIV hot spots.We employed Geodetector to measure the spatial stratifed heterogeneity(SSH)of HIV prevalence for adult women and men.We assessed the comprehensive correct knowledge(CCK)about HIV/AIDS and HIV testing uptake by employing the Least Absolute Shrinkage and Selection Operator(LASSO)regression to predict which combinations of CCKs can scale up the ratio of HIV testing uptake with sex-specifc needs.Results:In our analysis,we leveraged data for 158,408 respondents from 11 surveys in the MRU.From 2005–2015,Cote d’Ivoire was the hot spot for HIV prevalence with a Gi_Bin score of 3,Z-Score 8.0–10.1 and P<0.001.From 2016 to 2020,Guinea and Sierra Leone were hot spots for HIV prevalence with a Gi_Bin score of 2,Z-Score of 3.17 and P<0.01.The SSH confrmed the signifcant diferences in HIV prevalence at the national level strata,with a higher level for Cote d’Ivoire compared to other countries in both sexes with q-values of 0.61 and 0.40,respectively.Our LASSO model predicted diferent combinations of CCKs with sex-specifc needs to improve HIV testing uptake.Conclusions:The spatial distribution of HIV prevalence in the MRU is skewed and the CCK about HIV/AIDS and HIV testing uptake are far below the threshold target set by UNAIDS for ending the epidemic in the sub-region.Geodetector detected statistically signifcant SSH within and between countries in the MRU.Our LASSO model predicted that diferent emphases should be implemented when popularizing the CCK about HIV/AIDS for adult women and men.