检索结果-南通市图书馆

Estimatio. o. Finite Po.ulatio. To.als in High Dimensio.al Spaces

o.en Jo.rnal o. Statistics 2022年第5期12卷 581-596页

作者： Festus A. Were george o. orwa.roma.us o. otieno Department o. Statistics and Actuarial Sciences JKUAT Nairobi Kenya

In this paper, the pro.lem o. No.parametric Estimatio. o. Finite Po.ulatio. To.als in high dimensio.al datasets is co.sidered. A ro.ust estimato. o. the Finite Po.ulatio. To.al based o. Feedforward Backpro.agatio. Neural Netwo.k is derived with the aid o. a Super-Po.ulatio. Mo.el. This current study is mo.ivated by the fact that Lo.al Po.yno.ials and Kernel metho.s have in preceding related studies, been sho.n to.pro.ide go.d estimato.s fo. Finite Po.ulatio. To.als but in lo. dimensio.s. Even in these situatio.s ho.ever, bias at bo.ndary po.nts presents a big challenge when using these estimato.s in estimating Finite Po.ulatio. parameters. The challenge wo.sens as the dimensio. o. regresso.s increase. This is because as the dimensio. o. the Regresso. Vecto.s gro.s, the Sparseness o. the Regresso.s’ values in the design space beco.es unfeasible, resulting in a decrease in the fastest achievable rates o. co.vergence o. the Regressio. Functio. Estimato.s to.ards the target curve, rendering Kernel Metho.s and Lo.al Po.yno.ials ineffective to.address these challenges. This study co.siders the technique o. Artificial Neural Netwo.ks which yields ro.ust estimato.s in high dimensio.s and reduces the estimatio. bias with marginal increase in variance. This is due to.its Multi-Layer Structure, which can appro.imate a wide range o. functio.s to.any required level o. precisio.. The estimato.’s pro.erties are develo.ed, and a co.pariso. with existing estimato.s was co.ducted to.evaluate the estimato.’s perfo.mance using real data sets acquired fro. the United Natio.s Develo.ment Pro.ramme 2020. The estimatio. appro.ch perfo.ms well in an example using data fro. a United Natio.s Develo.ment Pro.ramme 2020 o. the study o. Human Develo.ment Index against o.her facto.s. The theo.etical and practical results imply that the Neural Netwo.k estimato. is highly reco.mended fo. survey sampling estimatio. o. the finite po.ulatio. to.al.

关键词： Neural Netwo.ks Kernel Smo.ther Lo.al Po.yno.ial No.parametric

来源：

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Ro.ust Classificatio. thro.gh a No.parametric Kernel Discriminant Analysis

引用

o.en Jo.rnal o. Statistics 2022年第4期12卷 443-455页

作者： Macdo.ald G. o.udho.george o. orwa.roma.us o. otieno Festus A. Were Department o. Statistics and Actuarial Sciences Jomo Kenyatta University of Agriculture and Technology Nairobi Kenya

The pro.lem o. classificatio. in situatio.s where the assumptio. o. no.mality in the data is vio.ated, and there are no.-linear clustered structures in the dataset is addressed. A ro.ust no.parametric kernel discriminant classificatio. functio., which is able to.address this challenge, has been develo.ed and the misclassificatio. rates co.puted fo. vario.s bandwidth matrices. A co.pariso. with existing parametric classificatio. functio.s such as the linear discriminant and quadratic discriminant is co.ducted to.evaluate the perfo.mance o. this classificatio. functio. using simulated datasets. The results presented in this paper sho. go.d perfo.mance in terms o. misclassificatio. rates fo. the kernel discriminant classifier when the co.rect bandwidth is selected as co.pared to.o.her identified existing classifiers. In this regard, the study reco.mends the use o. the pro.o.ed kernel discriminant classificatio. rule when o.e wishes to.classify units into.o.e o. several catego.ies o. po.ulatio. gro.ps where parametric classifiers might no. be applicable.

关键词： Discriminant Analysis Kernel Discriminant No.parametric

来源：

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Classificatio. o. Stateless Peo.le thro.gh a Ro.ust No.parametric Kernel Discriminant Functio.

引用

o.en Jo.rnal o. Statistics 2022年第5期12卷 563-580页

作者： Macdo.ald G. o.udho.george o. orwa.roma.us o. otieno Festus A. Were Department o. Statistics and Actuarial Sciences Jomo Kenyatta University of Agriculture and Technology Nairobi Kenya

Statelessness is the absence o. any Natio.ality. These include the Pemba, Sho.a, Galjeel, peo.le o. Burundi and Rwanda descent, and children bo.n in Kenya to.British o.erseas Citizens after 1983. Frequently, they are no. o.ly undo.umented but also.o.ten o.erlo.ked and no. included in Natio.al Administrative Registers. Acco.dingly, find it hard to.participate in So.ial and Eco.o.ic Affairs. There has been a majo. push by UNHCR and internatio.al partners to.“map” the size o. stateless po.ulatio.s and their demo.raphic pro.ile, as well as causes, po.ential so.utio.s and human rights situatio.. o.e o. the requirements by the UNHCR in their push is fo. co.ntries to.find a po.ential so.utio. to.statelessness which starts with classifying/asso.iating a perso. fro. these co.munities to.a particular lo.al co.munity that is reco.nized in Kenya. This paper addresses this pro.lem by ado.ting a Ro.ust No.parametric Kernel Discriminant functio. to.co.rectly classify the stateless co.munities in Kenya and co.pare the perfo.mance o. this metho. with the existing techniques thro.gh their classificatio. rates. This is because No.-parametric functio.s have pro.en to.be mo.e ro.ust and useful especially when there exists auxiliary info.matio. which can be used to.increase precisio.. The findings fro. this paper indicate that No.parametric discriminant classifiers pro.ide a go.d classificatio. metho. fo. classifying the stateless co.munities in Kenya. This is because they exhibit lo.er classificatio. rates co.pared to.the parametric metho.s such as Linear and Quadratic discriminant functio.s. In additio., the finding sho.s that based o. certain similarities in characteristics that exist in these co.munities that surro.nd the Pemba Co.munity, the Pemba co.munity can be classified as Giriama o. Rabai in which they seem to.have a stro.g link. In this regard, the study reco.mends the use o. the Kernel discriminant classifiers in classifying the stateless perso.s and that the Go.ernment o. K

关键词： Discriminant Analysis Kernel Discriminant No.parametric Classificatio. Statelessness

来源：

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Estimatio. o. Po.ulatio. Variance using the Co.fficient o. Kurto.is and Median o. an Auxiliary Variable under Simple Rando. Sampling

引用

o.en Jo.rnal o. Statistics 2017年第6期7卷 944-955页

作者： To.ui Kiplangat Milto. romanus otieno o.hiambo.george otieno orwa Pan African University Institute for Basic Sciences Technology and Innovation (PAUISTI) Nairobi Kenya Department o. Statistics and Actuarial Science Jomo Kenyatta University of Agriculture and Technology Nairobi Kenya

In this study we have pro.o.ed a mo.ified ratio.type estimato. fo. po.ulatio. variance o. the study variable y under simple rando. sampling witho.t replacement making use o. co.fficient o. kurto.is and median o. an auxiliary variable x. The estimato.’s pro.erties have been derived up to.first o.der o. Taylo.’s series expansio.. The efficiency co.ditio.s derived theo.etically under which the pro.o.ed estimato. perfo.ms better than existing estimato.s. Empirical studies have been do.e using real po.ulatio.s to.demo.strate the perfo.mance o. the develo.ed estimato. in co.pariso. with the existing estimato.s. The pro.o.ed estimato. as illustrated by the empirical studies perfo.ms better than the existing estimato.s under so.e specified co.ditio.s i.e. it has the smallest Mean Squared Erro. and the highest Percentage Relative Efficiency. The develo.ed estimato. therefo.e is suitable to.be applied to.situatio.s in which the variable o. interest has a po.itive co.relatio. with the auxiliary variable.

关键词： Mo.ified Ratio Type Variance Estimato. Study Variable Auxiliary Variable Kurto.is Median Bias Mean Squared Erro. (MSE) Percentage Relative Efficiency (PRE) Simple Rando. Sampling

来源：

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Bias Co.rectio. Technique fo. Estimating Quantiles o. Finite Po.ulatio.s under Simple Rando. Sampling witho.t Replacement

引用

o.en Jo.rnal o. Statistics 2021年第5期11卷 854-869页

作者： Nicho.as Makumi romanus o.hiambo.otieno george otieno orwa Festus Were Habineza Alexis Pan African University Institute for Basic Sciences Technology and Innovation Nairobi Kenya Department o. Mathematics Meru University of Science and Technology Meru Kenya Department o. Statistics and Actuarial Sciences JKUAT Nairobi Kenya

In this paper, the pro.lem o. no.parametric estimatio. o. finite po.ulatio. quantile functio. using multiplicative bias co.rectio. technique is co.sidered. A ro.ust estimato. o. the finite po.ulatio. quantile functio. based o. multiplicative bias co.rectio. is derived with the aid o. a super po.ulatio. mo.el. Mo.t studies have co.centrated o. kernel smo.thers in the estimatio. o. regressio. functio.s. This technique has also.been applied to.vario.s metho.s o. no.-parametric estimatio. o. the finite po.ulatio. quantile already under review. A majo. pro.lem with the use o. no.parametric kernel-based regressio. o.er a finite interval, such as the estimatio. o. finite po.ulatio. quantities, is bias at bo.ndary po.nts. By co.recting the bo.ndary pro.lems asso.iated with previo.s mo.el-based estimato.s, the multiplicative bias co.rected estimato. pro.uced better results in estimating the finite po.ulatio. quantile functio.. Furthermo.e, the asympto.ic behavio. o. the pro.o.ed estimato.s is presented. o.t-family:Verdana;">It is o.served that the estimato. is asympto.ically unbiased and statistically co.sistent when certain co.ditio.s are satisfied. The simulatio. results sho. that the suggested estimato. is quite well in terms o. relative bias, mean squared erro., and relative ro.t mean erro.. As a result, the multiplicative bias co.rected estimato. is stro.gly suggested fo. survey sampling estimatio. o. the finite po.ulatio. quantile functio..

关键词： Quantile Functio. Kernel Estimato. Multiplicative Bias Co.rectio. Technique Simple Rando. Sampling witho.t Replacement

来源：

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Indigeno.s Kno.ledge used in Breeding and Management o. Capra hircus Po.ulatio.s in Kajiado.and Makueni Co.nties, Kenya

引用

o.en Jo.rnal o. Genetics 2015年第3期5卷 111-135页

作者： o.ello.george otieno Jo.eph o.ino.Junga M. S. Badamana Jo.hua o. Amimo Department Name o. Animal pro.uctio. University of Nairobi Nairobi Kenya

The aim o. this pro.ect was to.study indigeno.s breeding practices used in management o. Capra hircus (Galla and Small East African go.ts) po.ulatio.s in Kajiado.Co.nty in Rift valley pro.ince and Makueni Co.nty in Eastern pro.ince. Data were o.tained thro.gh a field survey using questio.naires (Appendix I) and perso.al o.servatio.s between 1st September, 2013 and 2nd December, 2013. This study co.ered key characteristics o. go.ts pro.uctio. and areas o. go.t breeding, such as general farm details, number o. go.ts, main activities o. the farmers, farming types, breeds, flo.k structure, feeding, ho.sing, catastro.hes, selectio., mating systems, breeding systems, average age at parturitio., breeding pro.lems, and the impo.tance o. go.ts. Results sho.ed that the average number o. go.ts in Kajiado.was 100.65 ± std 49.88 while in Makueni it was 12.28 ± std 6.46. The main activity o. the peo.le interviewed was farming as 58 peo.le (96.67%) in Kajiado.and 42 peo.le (61.60%) cho.e farming as their main activity because this was their main so.rce o. liveliho.d. Flo.ks were do.inated by breeding females at a mean o. 39.06 ± std 16.75 in Kajiado.and a mean o. 5.62 ± std 3.50 in Makueni because females were kept to.repro.uce to.increase the size o. the flo.k and the males were kept majo.ly fo. cash and o.ly o.e o. two.were left to.repro.uce with the females. Dro.ght was the majo. catastro.he as it killed an average number o. go.ts o. 6.33 ± std 4.36. Pneumo.ia and diarrho.a were the majo. diseases acco.ding to.28 farmers (46.66%) in Kajiado.and 31 farmers (51.66%) in Makueni. Ticks and fleas were the majo. parasites acco.ding to.42 farmers (70%) in Kajiado.and 4 farmers (63.34%) in Makueni. Treatment was mo.tly do.e by the farmers individually as 54 farmers (90%) in Kajiado.and 46 farmers (76.67%) in Makueni treated the animals by themselves. This was so.because it was either no. easy to.get a veterinarian o. expensive fo. them to.hire veterinarian do.to.s. So.e farmers used tr

关键词： Indigeno.s Kno.ledge Breeding Management Capra Hircus Kajiado and Makueni Co.nties

来源：

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

o.e-Sample Bayesian Predictive Analyses fo. a No.ho.o.eneo.s Po.sso. Pro.ess with Delayed S-Shaped Intensity Functio. using No.-Info.mative Prio.s

引用

o.en Jo.rnal o. Statistics 2023年第5期13卷 717-733页

作者： otieno Co.lins o.awo.Luke Ako.g’o.Matiri george Munene Department o. Mathematics Egerton University Egerton Kenya

The delayed S-shaped so.tware reliability gro.th mo.el (SRGM) is o.e o. the no.-ho.o.eneo.s Po.sso. pro.ess (NHPP) mo.els which have been pro.o.ed fo. so.tware reliability assessment. The mo.el is distinctive because it has a mean value functio. that reflects the delay in failure repo.ting: there is a delay between failure detectio. and repo.ting time. The mo.el captures erro. detectio., iso.atio., and remo.al pro.esses, thus is appro.riate fo. so.tware reliability analysis. Predictive analysis in so.tware testing is useful in mo.ifying, debugging, and determining when to.terminate so.tware develo.ment testing pro.esses. Ho.ever, Bayesian predictive analyses o. the delayed S-shaped mo.el have no. been extensively explo.ed. This paper uses the delayed S-shaped SRGM to.address fo.r issues in o.e-sample predictio. asso.iated with the so.tware develo.ment testing pro.ess. Bayesian appro.ch based o. no.-info.mative prio.s was used to.derive explicit so.utio.s fo. the fo.r issues, and the develo.ed metho.o.o.ies were illustrated using real data.

关键词： Failure Intensity No.-Info.mative Prio.s So.tware Reliability Mo.el Bayesian Appro.ch No.-Ho.o.eneo.s Po.sso. Pro.ess

来源：

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Facto.s Affecting Investment Amo.gst Kenyan Fo.tballers： A Study o. Selected Fo.tballers in Kenya

引用

Eco.o.ics Wo.ld 2014年第6期2卷 369-377页

作者： o.aya o.car otieno george Misigah Caro.yne Kinyanjui Egerto. University Nakuru Kenya JKUAT Nakuru Kenya

The main o.jective o. the paper is to.examine the facto.s that influence investment initiatives amo.gst Kenyan fo.tballers who.are lo.ally based. Descriptive research design was used fo. the research. The po.ulatio. o. the study co.prised o. the players in the Kenyan premier league teams with players o. co.tract. Stratificatio. was do.e and each stratum a simple rando. sample was used to.select the sample o. 94 respo.dents. The research instrument fo. co.lectio. o. data was a questio.naire. The significance test was carried o.t using F-statistics to.co.pare the means in vario.s variables that yielded no.difference in the different variables tested. Co.fficient o. co.relatio. was used to.test the degrees o. co.relatio. o. the variables which generally resulted in all po.itive co.relatio.s but with different degrees. Generally, the findings were that there is a lo. financial literacy amo.g the Kenyan fo.tballers, peer and parental had great influence o. the investment attitude o. the fo.tballers. There is a need to.pro.ide extra kno.ledge o. financial matters to.the Kenyan fo.tball players, the Kenyan fo.tball clubs and the stakeho.ders. Effo.ts sho.ld be do.e to.increase fo.tball players＇ participatio. in the management o. perso.al finances. Immediate measure to.inco.po.ate all stakeho.ders and pro.essio.als is used to.initiate educatio. and training to.fo.tball players in Kenya o. perso.al financial matters to.help them invest in lo.g-term steady inco.e generating pro.ects.

关键词： financial literacy investment financial management

来源：

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Integratio. o. Gro.nd Magnetics and Energy Dispersive X-Ray Fluo.escence Spectro.co.y in Ilmenite Pro.pectio. in Magao.i, Kenya

引用

o.en Jo.rnal o. Applied Sciences 2016年第2期6卷 79-88页

作者： george o.uo. otieno Jo.n Gito.ga Githiri Willis Jakanyango.Ambusso Kenyatta University Nairobi Kenya Jo.o.Kenyatta University o. Agriculture Science and Technology Nairobi Kenya

The geo.o.y o. Magao.i area is asso.iated with the presence o. heavy minerals [1]. Magao.i’s neighbo.rs Maumba and Nguluku where ilmenite was disco.ered by Tio.in Reso.rce Inc. in 1996, using drilling and chemical analysis [2]. Ilmenite mineral is kno.n to.be magnetically weak, but pro.ides o.servable magnetic respo.se [3]. In this study, gro.nd magnetic survey metho. was carried o.t to.map magnetic ano.alies o. established statio.s, asso.iated with ilmenite bearing fo.matio.s. The magnetic co.to.r map plo.ted sho.ed weak and shallo. magnetic signatures spread thro.gho.t the study area. 2D Euler deco.vo.utio. so.utio.s revealed presence o. magnetised fo.matio.s fro. near surface to.a maximum depth o. abo.t 450 m at so.e po.nts. The weak magnetic fo.matio.s o. near surface indicated presence o. ilmenite. Energy dispersive X-ray spectro.co.y was do.e o. so.l samples co.lected rando.ly fro. the study area to.determine the percentage o. iro. and titanium o.ides. The results sho.ed elevated values o. titanium dio.ide, ranging fro. 1.5% to.13% which is way abo.e the glo.al average o. abo.t 0.7% [4]. The percentage o. iro. o.ide was lo., ranging fro. 1.5% to.4%, this being the reaso. fo. weak magnetisatio. o. the study area.

关键词： Ilmenite Magnetic Survey Euler Deco.vo.utio. Spectro.co.y

来源：

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Stem Bo.er Species Co.po.itio. o. Maize and Two.No.-Cereal Ho.ts in the Fo.est Zo.e o. Kisangani, DRC

引用

Jo.rnal o. Agricultural Science and Techno.o.y(A) 2014年第10期4卷 822-829页

作者： o.esime Mubenga Kanko.da Dudu Akaibe Walangululu Massamba george otieno o.g'amo.Bruno.Pierre Le Ru Labo.ato.y o. Eco.o.y and Management o. Animal Reso.rces Faculty of Science University of Kisangani P.O. Box 2012Kisangani Democratic Republic of Congo Faculty o. Agricultural Sciences Catholic University of Bukavu P.O. Box 285 Bukavu Democratic Republic of Congo Scho.l o. Bio.o.ical Science College of Physical and Biological Sciences (Chiromo Campus) University of Nairobi P.O. Box30197 Nairobi Kenya Institut de Recherche po.r le Do.elo.pement UR072 c/o ICIPE P.O. Box 30772 Nairobi Kenya and University Paris-Sud 11Orsay Cedex 91405 France

Lepido.teran stem bo.ers are the mo.t damaging pests o. maize in Sub-Saharan Africa. Despite the gro.ing impo.tance o. maize in the fo.est zo.e o. Demo.ratic Republic o. Co.go. no.data is available regarding stem bo.er pest species present and their relative impo.tance. It is thus impo.tant to.gather info.matio. likely to.guide future research in this area. This study was undertaken to.catalo.ue stem bo.er pest species identity and assess their relative infestatio. levels o. maize. Surveys were carried o.t in wild and cultivated habitats in Kisangani. Five species were co.lected o. maize, i.e., Sesamia calamistis Hampso. （1910）, Eldana saccharina Walker （1865）, Busseo.a fusca Fuller （1901）, Chilo.sp. Strand （1913）, and Mussidia nigrivenella Rago.o. （1888）. In the wild habitats, Po.no.a serrata Hampso., B. fusca and S. calamistis were co.lected o. Pennisetum purpureum whereas Chilo.sp. was co.lected o. Panicum maximum. o.r results suggest that P. maximum might affect the po.ulatio. dynamics o. Chilo.sp. whereas P. purpureum is expected no. to.influence the po.ulatio. dynamics o. o.her stem bo.ers o.ing to.its scarcity in the interio. o. the fo.est.

关键词： Stem bo.ers maize Busseo.afusca Chilo.sp. Eldana saccharina fo.est zo.e Kisangani.

来源：

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

欢迎您,

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

请选择保存的检索档案：

请选择收藏分类：

通借通还

欢迎您,

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：