Canonical correlation and discriminant analysis

Author(s):  
J. O. Ramsay ◽  
B. W. Silverman
1987 ◽  
Vol 64 (3) ◽  
pp. 823-827 ◽  
Author(s):  
Mark A. Brooks ◽  
Larry W. Boleach ◽  
J. L. Mayhew

To determine the predictive potential of selected cognitive and psychomotor variables to estimate basketball performance, 50 male high school players from 3 schools in the same conference were evaluated. One team won the Iowa state championship; the second team had a 12 and 10 record while the third team had a 4 and 16 record. The 3 coaches rated each player's ability from 1 to 10. Multiple regression analysis to predict coaches' rating of ability from vertical jump, hand reaction time, weight, and playing experience gave an R of .76. However, discriminant analysis to classify players on the 3 teams indicated as important knowledge about basketball, dribbling, shooting accuracy, and height. The canonical correlation for the 4 variables and team membership was .64. Using the 4 variables, 60% of the players could be correctly classified to their teams.


2016 ◽  
Author(s):  
Muhammad Yousefnezhad ◽  
Daoqiang Zhang

AbstractMultivariate Pattern (MVP) classification can map different cognitive states to the brain tasks. One of the main challenges in MVP analysis is validating the generated results across subjects. However, analyzing multi-subject fMRI data requires accurate functional alignments between neuronal activities of different subjects, which can rapidly increase the performance and robustness of the final results. Hyperalignment (HA) is one of the most effective functional alignment methods, which can be mathematically formulated by the Canonical Correlation Analysis (CCA) methods. Since HA mostly uses the unsupervised CCA techniques, its solution may not be optimized for MVP analysis. By incorporating the idea of Local Discriminant Analysis (LDA) into CCA, this paper proposes Local Discriminant Hyperalignment (LDHA) as a novel supervised HA method, which can provide better functional alignment for MVP analysis. Indeed, the locality is defined based on the stimuli categories in the train-set, where the correlation between all stimuli in the same category will be maximized and the correlation between distinct categories of stimuli approaches to near zero. Experimental studies on multi-subject MVP analysis confirm that the LDHA method achieves superior performance to other state-of-the-art HA algorithms.


2012 ◽  
Vol 16 (3) ◽  
pp. 213-236 ◽  
Author(s):  
Agnieszka Szymańska

Abstract The research aimed to describe the differences in preschool children’s families in terms of parental influence. One of the important factors predicting a child’s behavior at kindergarten was parental directiveness. Directiveness was conceptualized as one of the acts of speech by which the speaker coaxes another to do something. Two types of directiveness were distinguished: warm-hearted directiveness and aggressive directiveness. Two hundred and four participants, parents of kindergarten children, took part in the research. Selection for the research sample was conducted according to the teacher’s representation of a child’s behavior at kindergarten (well-behaved or badly-behaved). Parents completed psychological tests measuring their level of parental control (conceptualized as teaching the child the rules of social behavior) and, finally, the level and type of directiveness (warm-hearted or aggressive). The purpose of the analysis was to discover which of the enumerated variables best explained a child’s behavior at kindergarten. Canonical correlation, discriminant analysis and data mining methods were used for the analysis. Analyses were performed with the help of the Statistical Package for Social Sciences (SPSS) and STATISTICA Data Miner 8 software. The results indicate that the level and type of parental directiveness is the most important factor that distinguishes children in groups, the split being due to the children’s behavior at kindergarten.


2020 ◽  
Vol 22 (4) ◽  
pp. 282-291
Author(s):  
Revi Rosavika Kinansi ◽  
Mega Tyas Prihatin

Discriminant analysis is one of the statistical techniques that may use to provide the most appropriate estimation for classifying individuals into one group based on the independent variable score (discriminant score). There are 2 main assumptions in discriminant analysis such as fulfilled data normality and similarity of variant-covariants. This study aims to determine whether there is a relationship between DHF Incidence Rate (IR) and entomology index if a region is classified as a coast-not a coast and rural-urban. This research conducted in 78 districts in Indonesia carried out in Disease Reservoir and Vector Specific Research from 2016 to 2017. The geographical area of ​​Indonesia which has a tropical climate with three months of rainy season in December, January, February and three months of the dry season in June, July, August can be a hyperendemic area of ​​DHF. This condition is exacerbated by the development of increasingly complex urban areas and the development of rural areas into cities that reduce environmental quality and have an impact on the expansion of the habitat of Aedes aegypti as vector of DHF. The data to be analyzed are the entomology index in the form of numbers of HI, BI, CI and ABJ against IR. The results of the analysis provide information that the very low value of Canonical Correlation is 0.076 classified as coast and not coast so that there is no relationship between the independent variable and the dependent variable. While the Canonical Correlation value is quite high, which is 0.219 classified as rural and urban showed that there is a relationship between the independent variable and the dependent variable. Based on the results, densely populated ecosystems in urban or rural areas have a great chance of cases of dengue hemorrhagic fever, so people need to monitor mosquito larvae to control DHF. Abstrak Analisis diskriminan adalah salah satu teknik statistik yang dapat digunakan untuk memberikan pendugaan yang paling tepat untuk mengklasifikasikan individu ke dalam salah satu kelompok berdasarkan skor variabel bebas (skor diskriminan). Terdapat 2 asumsi utama dalam melakukan analisis diskriminan, yaitu normalitas data harus terpenuhi dan kesamaan varian-kovarian. Penelitian ini bertujuan untuk mengetahui apakah terdapat hubungan antara Incidence Rate (IR) DBD dengan indeks entomologi jika suatu wilayah diklasifikasi menjadi pantai-bukan pantai dan perdesaan-perkotaan. Penelitian telah dilakukan di 78 kabupaten di Indonesia pada Riset Khusus Vektor dan Reservoir Penyakit tahun 2016 hingga 2017. Wilayah geografis Indonesia yang beriklim tropis dengan tiga bulan musim hujan pada Desember, Januari, Februari dan tiga bulan musim kemarau pada Juni, Juli, Agustus dapat menjadi wilayah hiperendemik DBD. Kondisi tersebut diperparah oleh perkembangan wilayah perkotaan yang semakin kompleks dan perkembangan wilayah pedesaan menjadi kota yang menurunkan kualitas lingkungan hidup dan berdampak pada perluasan habitat nyamuk Aedes aegypti vektor penyakit DBD. Data yang akan dianalisis adalah data indeks entomologi berupa angka HI, BI, CI dan ABJ terhadap IR. Hasil analisis memberikan informasi bahwa nilai Canonical Correlation yang sangat rendah yaitu 0,076, jika diklasifikasi menjadi pantai dan bukan pantai menunjukkan tidak terdapat hubungan antara variabel bebas dengan variabel terikat. Nilai Canonical Correlation yang cukup tinggi yaitu 0,219, jika diklasifikasi menjadi perdesaan dan perkotaan menunjukkan terdapat hubungan antara variabel bebas dengan variabel terikat nya. Berdasarkan hasil penelitian ini, ekosistem padat penduduk di perkotaan atau perdesaan memiliki peluang besar terhadap adanya kasus demam berdarah dengue, sehingga masyarakat perlu melakukan monitoring terhadap jentik nyamuk untuk pengendalian DBD.


INFERENSI ◽  
2013 ◽  
Vol 7 (1) ◽  
pp. 143
Author(s):  
Suryani Suryani

This study attempts to determine the effect of customer satisfaction in the context of the marketing mixof positive word of mouth customers of PT. Bank Muamalat Indonesia, Tbk Medan Branch. Number of questionnaires was collected as many as 50 questionnaires respondent. Analysis technique used was factor analysis followed by discriminant analysis with program of SPSS version 17.0. Based on factor analysis, it is known that one factor called location was proved to be inaccurate (correlation value <0.5) so that it is excluded from the model. Canonical correlation is obtained for 0.593 with a significance of 0.003. It can be concluded that 35.20% of the variation of positive word of mouth can be explained by the variable of discriminant Product, Promotion, Price, People, Physics and Process.


Sign in / Sign up

Export Citation Format

Share Document