Improving the Effectiveness of Image Classification Structural Methods by Compressing the Description According to the Information Content Criterion

The research aims to improve the performance of image recognition methods based on a description in the form of a set of keypoint descriptors.The main focus is on increasing the speed of establishing the relevance of object and etalon descriptions while maintaining the required level of classification efficiency.The class to be recognized is represented by an infinite set of images obtained from the etalon by applying arbitrary geometric transformations.It is proposed to reduce the descriptions for the etalon database by selecting the most significant descriptor components according to the information content criterion.The informativeness of an etalon descriptor is estimated by the difference of the closest distances to its own and other descriptions.The developed method determines the relevance of the full description of the recognized object with the reduced description of the etalons.Several practical models of the classifier with different options for establishing the correspondence between object descriptors and etalons are considered.The results of the experimental modeling of the proposed methods for a database including images of museum jewelry are presented.The test sample is formed as a set of images from the etalon database and out of the database with the application of geometric transformations of scale and rotation in the field of view.The practical problems of determining the threshold for the number of votes,based on which a classification decision is made,have been researched.Modeling has revealed the practical possibility of tenfold reducing descriptions with full preservation of classification accuracy.Reducing the descriptions by twenty times in the experiment leads to slightly decreased accuracy.The speed of the analysis increases in proportion to the degree of reduction.The use of reduction by the informativeness criterion confirmed the possibility of obtaining the most significant subset of features for classification,which guarantees a decent level of accuracy.

Correlation of preoperative inflammatory factors and emotional disorders with postoperative delirium in patients with craniocerebral trauma

BACKGROUND Traumatic brain injury(TBI)imposes a substantial societal and familial burden due to its high disability and fatality rates,rendering it a serious public health problem.Some patients with TBI have poor treatment outcomes and are prone to postoperative delirium(POD),which affects their quality of life.Anxiety has been linked to increased POD incidence in some studies,while others have found no correlation.AIM To investigate the correlation of POD risk factors,preoperative inflammatory factors,and mood disorders in patients with TBI.METHODS We retrospectively collected data on the treatment of 80 patients with TBI from November 2021 to September 2023.Patients were grouped as POD and non-POD,according to their POD status,and the general data of the two groups were compared.Inflammatory factor levels were detected preoperatively,and the Hamilton Depression Scale(HAMD)and Hamilton Anxiety Scale(HAMA)were used to investigate the risk factors associated with POD in these patients.Logistic regression was used to identify the independent risk factors.RESULTS Twenty-one patients(26.25%)developed POD,including 7,10,and 4 cases of the excitatory,inhibitory,and mixed types,respectively.There were 59 cases(73.75%)in the non-POD group.Compared with the non-POD group,the POD group had a significantly higher proportion of patients with low Glasgow Coma Scale(GCS)scores before admission,unilateral mydriasis,preoperative hemorrhagic shock,intraventricular hemorrhage(IVH),and postoperative hyperglycemic hyperosmolar disease(P<0.05).In the POD group,interleukin-6(IL-6),human tumor necrosis factor-α(TNF-α),myeloperoxidase levels,HAMA,and HAMD scores were higher than those in the non-POD group(all P<0.05).Logistic multivariate analysis showed that GCS score at admission,IVH,IL-6,TNF-α,HAMA,and HAMD were independent risk factors for POD in patients with TBI(P<0.05).CONCLUSION Low GCS score at admission,IVH,elevated IL-6 and TNF-α,other inflammatory indicators,anxiety,and depression,can increase the risk of POD in patients with TBI after surgery.

Correlation between psychological traits and the use of smart medical services in young and middle-aged adults: An observational study

BACKGROUND Psychological problems affect economic development.However,there is a huge gap between mental health service resources and mental health service needs.Existing mental health service technology and platforms cannot meet all the diverse mental health needs of people.Smart medicine is a new medical system based online that can effectively improve the quality and efficiency of medical services and make mental health services accessible.AIM To explore the level of intelligent medical use among young and middle-aged people and its correlation with psychological factors.METHODS Convenience sampling was used to select 200 young and middle-aged patients with medical experience at the Third People's Hospital of Chengdu between January 2022 and January 2023 as the research subjects.The general condition Questionnaire,Eysenck Personality Questionnaire,Symptom Checklist-90,General Health Questionnaire,and Smart Medical Service Use Intention Questionnaire were used to collect data.Pearson’s correlation was used to analyze the correlation between the participants’willingness to use smart medical services and their personality characteristics,psychological symptoms,and mental health.RESULTS The results revealed that the mental health of young and middle-aged people was poor,and some had psycho-logical problems such as anxiety,depression,and physical discomfort.Familiarity,acceptance,and usage of smart healthcare in this population are at a medium level,and these levels correlate with psychological characteristics.Acceptance was positively correlated with E,and negatively correlated with P,anxiety,fear,anxiety/insomnia,and social dysfunction.The degree of use was negatively correlated with P,obsessive-compulsive symptoms,depression,anxiety,hostility,paranoia,and somatic symptoms.CONCLUSION The familiarity,acceptance,and usage of smart medical services among the middle-aged and young groups are related to various psychological characteristics.

High rate of clinically relevant improvement following anatomical total shoulder arthroplasty for glenohumeral osteoarthritis

BACKGROUND The minimal clinically important difference(MCID)is defined as the smallest meaningful change in a health domain that a patient would identify as important.Thus,an improvement that exceeds the MCID can be used to define a successful treatment for the individual patient.AIM To quantify the rate of clinical improvement following anatomical total shoulder arthroplasty for glenohumeral osteoarthritis.METHODS Patients were treated with the Global Unite total shoulder platform arthroplasty between March 2017 and February 2019 at Herlev and Gentofte Hospital,Denmark.The patients were evaluated preoperatively and 3 months,6 months,12 months,and 24 months postoperatively using the Western Ontario Osteoarthritis of the Shoulder index(WOOS),Oxford Shoulder Score(OSS)and Constant-Murley Score(CMS).The rate of clinically relevant improvement was defined as the proportion of patients who had an improvement 24 months postoperatively that exceeded the MCID.Based on previous literature,MCID for WOOS,OSS,and CMS were defined as 12.3,4.3,and 12.8 respectively.RESULTS Forty-nine patients with a Global Unite total shoulder platform arthroplasty were included for the final analysis.Mean age at the time of surgery was 66 years(range 49.0-79.0,SD:8.3)and 65%were women.One patient was revised within the two years follow-up.The mean improvement from the preoperative assessment to the two-year follow-up was 46.1 points[95%confidence interval(95%CI):39.7-53.3,P<0.005]for WOOS,18.2 points(95%CI:15.5-21.0,P<0.005)for OSS and 37.8 points(95%CI:31.5-44.0,P<0.005)for CMS.Two years postoperatively,41 patients(87%)had an improvement in WOOS that exceeded the MCID,45 patients(94%)had an improvement in OSS that exceeded the MCID,and 42 patients(88%)had an improvement in CMS that exceeded the MCID.CONCLUSION Based on three shoulder-specific outcome measures we find that approximately 90%of patients has a clinically relevant improvement.This is a clear message when informing patients about their prognosis.

Contribution and Relevance of Transfontanellar Ultrasound in Newborns in Ouagadougou: Multicentre Study on 1000 Cases

Objectives: To demonstrate the contribution and relevance of ETFs through the study of 1000 examination reports carried out in the medical imaging departments of the OUAGADOUGOU CHU. Material and method: Analytical descriptive study with retrospective collection, extended from 1st January 2020 to 1st January 2022. Results: Of the 1000 transfontanellar ultrasound reports, the mean age of patients was 7.61 +/ 7.5 days, with extremes of zero and 28 days. Sex was specified in 989 cases. Males accounted for 54.49% and females for 45.51%. 555 transfontanellar ultrasound were performed in 2020. 441 in 2021 and 4 in 2022. 61.9% of transfontanellar ultrasound were performed at the Bogodogo University Hospital, 205 at Charles de Gaulle and 176 at Tengandogo. Indications for transfontanellar ultrasound were dominated by neonatal distress (65.8%), followed by convulsions (10.2%) and prematurity (9.1%). Transfontanellar ultrasound was normal in 570 cases (57%) and abnormal in 430 cases (43%). Abnormalities were dominated by haemorrhage and ischaemic lesions in 66.28% (285) and 21.63% (93) of cases respectively. In the group of normal transfontanellar ultrasound, neonatal distress represented 59.65% of indications and prematurity 10.7% of indications. As for abnormal transfontanellar ultrasound, neonatal suffering accounted for 73.95% of indications and convulsions for 12.56%. The average age ofpatients with an abnormal transfontanellar ultrasound was 8.74 days +/ 7.89 days. The indication for investigations was relevant in 42.2% of cases and irrelevant in 57.8%;of the transfontanellar ultrasound with relevant indications, 0.71 were normal and 99.29 abnormal;of the transfontanellar ultrasound with irrelevant indications, the transfontanellar ultrasound was normal in 98.1% and abnormal in 1.9%. Conclusion: Transfontanellar ultrasound is an important part of ultrasound in current practice. Haemorrhage, anoxic-ischaemic lesions and hydrocephalus are the most frequent pathologies found by this technique in newborns. Whether or not the examinations were normal depended on the appropriateness of the prescription.

Application Analysis of Nursing Students'Grades in Course Relevance Based on Association Rule Mining Algorithm Apriori

By analyzing the correlation between courses in students’grades,we can provide a decision-making basis for the revision of courses and syllabi,rationally optimize courses,and further improve teaching effects.With the help of IBM SPSS Modeler data mining software,this paper uses Apriori algorithm for association rule mining to conduct an in-depth analysis of the grades of nursing students in Shandong College of Traditional Chinese Medicine,and to explore the correlation between professional basic courses and professional core courses.Lastly,according to the detailed analysis of the mining results,valuable curriculum information will be found from the actual teaching data.

Handily etching nickel foams into catalyst-substrate fusion self‐stabilized electrodes toward industrial‐level water electrolysis

The key challenge of industrial water electrolysis is to design catalytic electrodes that can stabilize high current density with low power consumption(i.e.,overpotential),while industrial harsh conditions make the balance between electrode activity and stability more difficult.Here,we develop an efficient and durable electrode for water oxidation reaction(WOR),which yields a high current density of 1000 mA cm−2 at an overpotential of only 284 mV in 1M KOH at 25°C and shows robust stability even in 6M KOH strong alkali with an elevated temperature up to 80°C.This electrode is fabricated from a cheap nickel foam(NF)substrate through a simple one-step solution etching method,resulting in the growth of ultrafine phosphorus doped nickel-iron(oxy)hydroxide[P-(Ni,Fe)O_(x)H_(y)]nanoparticles embedded into abundant micropores on the surface,featured as a self-stabilized catalyst–substrate fusion electrode.Such self-stabilizing effect fastens highly active P-(Ni,Fe)O_(x)H_(y)species on conductive NF substrates with significant contribution to catalyst fixation and charge transfer,realizing a win–win tactics for WOR activity and durability at high current densities in harsh environments.This work affords a cost-effective WOR electrode that can well work at large current densities,suggestive of the rational design of catalyst electrodes toward industrial-scale water electrolysis.

Relevant Visual Semantic Context-Aware Attention-Based Dialog

The existing dataset for visual dialog comprises multiple rounds of questions and a diverse range of image contents.However,it faces challenges in overcoming visual semantic limitations,particularly in obtaining sufficient context from visual and textual aspects of images.This paper proposes a new visual dialog dataset called Diverse History-Dialog(DS-Dialog)to address the visual semantic limitations faced by the existing dataset.DS-Dialog groups relevant histories based on their respective Microsoft Common Objects in Context(MSCOCO)image categories and consolidates them for each image.Specifically,each MSCOCO image category consists of top relevant histories extracted based on their semantic relationships between the original image caption and historical context.These relevant histories are consolidated for each image,and DS-Dialog enhances the current dataset by adding new context-aware relevant history to provide more visual semantic context for each image.The new dataset is generated through several stages,including image semantic feature extraction,keyphrase extraction,relevant question extraction,and relevant history dialog generation.The DS-Dialog dataset contains about 2.6 million question-answer pairs,where 1.3 million pairs correspond to existing VisDial’s question-answer pairs,and the remaining 1.3 million pairs include a maximum of 5 image features for each VisDial image,with each image comprising 10-round relevant question-answer pairs.Moreover,a novel adaptive relevant history selection is proposed to resolve missing visual semantic information for each image.DS-Dialog is used to benchmark the performance of previous visual dialog models and achieves better performance than previous models.Specifically,the proposed DSDialog model achieves an 8% higher mean reciprocal rank(MRR),11% higher R@1%,6% higher R@5%,5% higher R@10%,and 8% higher normalized discounted cumulative gain(NDCG)compared to LF.DS-Dialog also achieves approximately 1 point improvement on R@k,mean,MRR,and NDCG compared to the original RVA,and 2 points improvement compared to LF andDualVD.These results demonstrates the importance of the relevant semantic historical context in enhancing the visual semantic relationship between textual and visual representations of the images and questions.

Research on the optimization strategy of customers’electricity consumption based on big data

Current power systems face significant challenges in supporting large-scale access to new energy sources,and the potential of existing flexible resources needs to be fully explored from the power supply,grid,and customer perspectives.This paper proposes a multi-objective electricity consumption optimization strategy considering the correlation between equipment and electricity consumption.It constructs a multi-objective electricity consumption optimization model that considers the correlation between equipment and electricity consumption to maximize economy and comfort.The results show that the proposed method can accurately assess the potential for electricity consumption optimization and obtain an optimal multi-objective electricity consumption strategy based on customers’actual electricity consumption demand.

Precise Multi-Class Classification of Brain Tumor via Optimization Based Relevance Vector Machine

The objective of this research is to examine the use of feature selection and classification methods for distinguishing different types of brain tumors.The brain tumor is characterized by an anomalous proliferation of brain cells that can either be benign or malignant.Most tumors are misdiagnosed due to the variabil-ity and complexity of lesions,which reduces the survival rate in patients.Diagno-sis of brain tumors via computer vision algorithms is a challenging task.Segmentation and classification of brain tumors are currently one of the most essential surgical and pharmaceutical procedures.Traditional brain tumor identi-fication techniques require manual segmentation or handcrafted feature extraction that is error-prone and time-consuming.Hence the proposed research work is mainly focused on medical image processing,which takes Magnetic Resonance Imaging(MRI)images as input and performs preprocessing,segmentation,fea-ture extraction,feature selection,similarity measurement,and classification steps for identifying brain tumors.Initially,the medianfilter is practically applied to the input image to reduce the noise.The graph-cut segmentation technique is used to segment the tumor region.The texture feature is extracted from the output of the segmented image.The extracted feature is selected by using the Ant Colony Opti-mization(ACO)algorithm to improve the performance of the classifier.This prob-abilistic approach is used to solve computing issues.The Euclidean distance is used to calculate the degree of similarity for each extracted feature.The selected feature value is given to the Relevance Vector Machine(RVM)which is a multi-class classification technique.Finally,the tumor is classified as abnormal or nor-mal.The experimental result reveals that the proposed RVM technique gives a better accuracy range of 98.87%when compared to the traditional Support Vector Machine(SVM)technique.

Typos Correction in Overseas Chinese Learning Based on Chinese Character Semantic Knowledge Graph

In recent years, more and more foreigners begin to learn Chinese characters, but they often make typos when using Chinese. The fundamental reason is that they mainly learn Chinese characters from the glyph and pronunciation, but do not master the semantics of Chinese characters. If they can understand the meaning of Chinese characters and form knowledge groups of the characters with relevant meanings, it can effectively improve learning efficiency. We achieve this goal by building a Chinese character semantic knowledge graph (CCSKG). In the process of building the knowledge graph, the semantic computing capacity of HowNet was utilized, and 104,187 associated edges were finally established for 6752 Chinese characters. Thanks to the development of deep learning, OpenHowNet releases the core data of HowNet and provides useful APIs for calculating the similarity between two words based on sememes. Therefore our method combines the advantages of data-driven and knowledge-driven. The proposed method treats Chinese sentences as subgraphs of the CCSKG and uses graph algorithms to correct Chinese typos and achieve good results. The experimental results show that compared with keras-bert and pycorrector + ernie, our method reduces the false acceptance rate by 38.28% and improves the recall rate by 40.91% in the field of learning Chinese as a foreign language. The CCSKG can help to promote Chinese overseas communication and international education.

Sustainability Evaluation of Mangrove Forest Management System of Tagbanua Tribe in Bgy. Manalo, Puerto Princesa City, Palawan, Philippines

Community-based forest management agreement in the country is a needed instrument in attaining sustainability of mangrove management.Sadly,there is no assurance that the system implemented in the mangrove forest management is sustainable.So,evaluating the mangrove management sustainability of the local tribe is a viable avenue for the appropriate management.In this study,the sustainability of the mangrove management system of the Tagbanua tribe in Bgy.Manalo,Puerto Princesa City,Palawan was evaluated.The study utilized various criteria with relevant indicators of sustainable mangrove forest management in assessing the mangrove forest management system.Focused group discussions were conducted to identify the relevant sustainable mangrove forest management C&I and verifiers.Each indicator was rated using the formulated verifiers in the form of the rating scale.Through household interviews,FGD,KII,mangrove assessment,and secondary data analysis,this study also used a mathematical model on the Sustainability Index for Individual Criteria(SIIC)to evaluate the scores for individual criteria and the Overall Sustainability Index(OSI)of the community.As a result,there are a total of seven relevant criteria,and 35 relevant indicators for Mangrove Management in Barangay Manalo.Based on the individual rating of seven criteria,the overall rating of the sustainable mangrove management system is 1.80,which implies a fairly sustainable mangrove management system.Also,the computed overall sustainability index is 0.26,which is fairly or moderately sustainable.Each criterion has strengths and weaknesses and needs to be improved to have a highly sustainable mangrove management system.

On Topics in Quantum Games

This work concentrates on simultaneous move non-cooperating quantum games. Part of it is evidently not new, but it is included for the sake self consistence, as it is devoted to introduction of the mathematical and physical grounds of the pertinent topics, and the way in which a simple classical game is modified to become a quantum game (a procedure referred to as a quantization of a classical game). The connection between game theory and information science is briefly stressed, and the role of quantum entanglement (that plays a central role in the theory of quantum games), is exposed. Armed with these tools, we investigate some basic concepts like the existence (or absence) of a pure strategy and mixed strategy Nash equilibrium and its relation with the degree of entanglement. The main results of this work are as follows: 1) Construction of a numerical algorithm based on the method of best response functions, designed to search for pure strategy Nash equilibrium in quantum games. The formalism is based on the discretization of a continuous variable into a mesh of points, and can be applied to quantum games that are built upon two-players two-strategies classical games, based on the method of best response functions. 2) Application of this algorithm to study the question of how the existence of pure strategy Nash equilibrium is related to the degree of entanglement (specified by a continuous parameter γ ). It is shown that when the classical game GC has a pure strategy Nash equilibrium that is not Pareto efficient, then the quantum game GQ with maximal entanglement (γ = π/2) has no pure strategy Nash equilibrium. By studying a non-symmetric prisoner dilemma game, it is found that there is a critical value 0γc such that for γγc there is a pure strategy Nash equilibrium and for γ≥γc there is no pure strategy Nash equilibrium. The behavior of the two payoffs as function of γ starts at that of the classical ones at (D, D) and approaches the cooperative classical ones at (C, C) (C = confess, D = don’t confess). 3) We then study Bayesian quantum games and show that under certain conditions, there is a pure strategy Nash equilibrium in such games even when entanglement is maximal. 4) We define the basic ingredients of a quantum game based on a two-player three strategies classical game. This requires the introduction of trits (instead of bits) and quantum trits (instead of quantum bits). It is proved that in this quantum game, there is no classical commensurability in the sense that the classical strategies are not obtained as a special case of the quantum strategies.

基于中红外光谱分析技术的香菇产地识别研究

采用中红外光谱分析技术对香菇产地进行识别研究，并将相关向量机（relevance vector machine ， RVM ）算法应用于中红外光谱判别分析之中，取得了较好的效果。通过采集香菇粉末的中红外透射光谱，去除光谱噪声明显部分，对剩下的3581～689 cm-1透射谱线采用多元散射校正（multiplicative scatter correc-tion ，MSC）进行预处理，并基于预处理谱线建立了香菇产地识别的偏最小二乘判别分析（partial least squares-discriminant analysis ，PLS-DA）、簇类独立软模式分类（soft independent modeling of class analogy ， SIMCA）、K 最邻近算法（K-nearest neighbor algorithm ， KNN ）、支持向量机（support vector machine ， SVM ）、RVM模型等五种判别分析模型。所有模型的识别正确率均高于80％，KNN ，SVM 和RVM 判别分析模型取得了相近的结果，建模集和预测集识别正确率高于90％。基于全谱的PLS-DA模型的加权回归系数，利用加权回归系数法选取了6个特征波数，并基于特征波数建立了PLS-DA ，KNN ，SVM 和RVM 模型。基于特征波数的PLS-DA模型的建模集和预测集识别正确率均低于80％，而KNN ，SVM和RVM模型的建模集和预测集的识别效果相近，且都高于90％。基于全谱和特征波数的模型中，RV M 算法表现出较好的效果，识别正确率优于90％。结果表明，基于中红外光谱技术能用于香菇产地的识别，特征波数的选择以及RVM算法可以有效的用于中红外光谱判别分析中。本文成功将中红外光谱用于香菇产地识别研究，为香菇品质以及其他农产品品质分析提供了一种新的想法，具有实际意义。

一种基于QPSO-RVM的模拟电路故障预测方法

提出了一种可应用于模拟电路故障预测的方法。通过提取被测电路的频域响应信号,计算皮尔逊相关系数,从而表征电路元件的健康度;在获取元件在不同时间点的健康度数据的基础上,推导出电路元件发生故障时的健康度阈值;将经量子粒子群算法优化的相关向量机算法用于故障预测,预测各个时间点的元件健康度变化轨迹并估计模拟电路的剩余有用寿命。该预测方法计算简单、通用性强,适用于实时预测。故障预测仿真实验与实例实验证明了方法的有效性与先进性。

基于综合健康指数与RVM的系统级失效预测

针对具有多维状态变量、多种工作模式和故障模式的复杂工程系统,提出一种基于综合健康指数(synthesized health index,SHI)与相关向量机(relevance vector machine,RVM)的系统级失效预测方法。在离线训练阶段,先根据有限失效历史数据建立各工作模式下的健康评估模型,并据此获得各历史退化轨迹的SHI序列;然后再使用RVM对这些序列进行回归处理,进而辨识出与回归曲线最为匹配的函数模型。在线预测阶段,先运用健康评估模型计算当前设备的SHI序列并进行RVM回归,再拟合出离线阶段确定的函数模型并添加时变噪声;最后,外推预测出系统剩余使用寿命的概率密度分布。该方法成功应用到涡轮发动机的失效预测案例。

Method of acquiring web features and its application in web search

Focusing on the problem that it is hard to utilize the web multi-fields information with various forms in large scale web search,a novel approach,which can automatically acquire features from web pages based on a set of well defined rules,is proposed.The features describe the contents of web pages from different aspects and they can be used to improve the ranking performance for web search.The acquired feature has the advantages of unified form and less noise,and can easily be used in web page relevance ranking.A special specs for judging the relevance between user queries and acquired features is also proposed.Experimental results show that the features acquired by the proposed approach and the feature relevance specs can significantly improve the relevance ranking performance for web search.

德里达翻译思想:回归、转变还是其他?

雅克·德里达将解构思想应用于翻译研究,形成了解构的翻译理论。解构翻译理论主要体现在德里达的一篇演讲稿"Quest-ce que-unetraduction ‘relevante’?"中,文中提出了一种"Relevant翻译"理论。本文回顾了德里达的解构翻译思想,重点剖析了"Relevant翻译"理论,并针对国内翻译界对德里达这一思想的两种解读,尤其是其中的一种误解,提出不同看法,说明德里达的"Relevant翻译"理论及其解构翻译思想与哲学史上其他一些思想之间的关系不是"回归"而是传承,这一理论也更不是德里达本人解构思想的一种"转变",而只是其解构思想在翻译理论中的深入和延续。

The Teachers＇ Feedback Act Research of College English Classes

The Follow up Move Research has been deepening after it was first defined in 1975 by Sinclair and Coulthard, but it has not been paid attention in China, let alone study its effect in the classroom. In this essay based on the author＇s literature reading, the Follow up move functions and their language features are classified; its pragmatic motivations are researched and some factors which affect its relevance of occurrence are studied. Then with this guide of the frame four teachers classroom follow up moves are studied. Finally, the other scholars＇ findings in this field are commented and then the author＇s own insights are put forward.

Estimation of illumination chromaticity via adaptive reduced relevance vector machine

A new regression algorithm of an adaptive reduced relevance vector machine is proposed to estimate the illumination chromaticity of an image for the purpose of color constancy. Within the framework of sparse Bayesian learning, the algorithm extends the relevance vector machine by combining global and local kernels adaptively in the form of multiple kernels, and the improved locality preserving projection （LLP） is then applied to reduce the column dimension of the multiple kernel input matrix to achieve less training time. To estimate the illumination chromaticity, the algorithm is trained by fuzzy central values of chromaticity histograms of a set of images and the corresponding illuminants. Experiments with real images indicate that the proposed algorithm performs better than the support vector machine and the relevance vector machine while requiring less training time than the relevance vector machine.