Master in Data Science (SMS, TU) Statistical Computing with R Question Paper 2082 (Set pages 13-14; Second Assessment 2082) Nepal

Describe the following concepts with examples: a) Wilkinson's approach of grammar of graphics b) ggplot2 approach of grammar of graphics

Discuss following concepts with clear examples: a) BLUE in linear regression b) LINE in linear regression

Explain followings concepts with clear examples: a) Spearman's rank correlation b) Mann-Whitney U test

Explain supervised learning classification regression model with: a) Multinominal logistic regression b) Random forest model

Discuss supervised linear regression models with focus on: a) Support vector machine b) Multilayer perceptron

Do the following in R Studio using ggplot2 and dplyr packages and knit output as PDF file: a) Create a dataset with following variables: age (20-59 years), height (110 – 190 centimeters), weight (40-90 kg) with random 150 cases of each variable. Your roll number must be used to set the random seed. b) Compute body mass index (BMI) variable as: BMI = [(weight in kg) / (height in meter squared)] c) Create body mass index categories: <18, 18-24, 25-30, 30+ and label them as "underweight", "normal", "overweight" and "obese" respectively using dplyr package d) Show the percentage distribution of labelled BMI variable with pie chart using ggplot2 package

Do the following in R Studio using "airquality" dataset of R with R script to knit PDF output: a) Perform test of normality of "Temp" variable by each category of "Month" variable and interpret all the results carefully b) Perform test of equality of variance on "Temp" variable by each category of Month variable and interpret the result carefully c) Compare Temp by Month using the best statistical test for this data and interpret the result carefully d) Perform the post-hoc test and interpret the results carefully

Do the following in R Studio with R script to knit PDF output: a) Create a dataset with 200 random cases, 1 random binary (1 and 0) variable and four random non-binary (categorical and continuous) variables with your roll number as random seed b) Divide it into train and test datasets with 70:30 random splits c) Fit a supervised logistic regression and decision tree classification models on train data with binary variable as dependent variable and all other four variables as independent variable d) Predict the released variable in the test datasets of both the models and interpret the results carefully

Do as follows using in-built "USArrests" dataset with R script to knit PDF output: a) Create a "criminality scale" of four variables of this dataset using the Principal Component Analysis b) Compute the eigenvalues and interpret the PCA result carefully using Kaiser's criteria c) Show the Scree plot and decide on the number of components to retain with careful interpretation d) Revise the criminality scale using VARIMAX rotation and interpret the result carefully

OR

Do as follows using given dataset of 10 US cities in R studio with R script:

a) Get dissimilarity distance as city.dissimilarity object b) Fit a classical multidimensional model using the city.dissimilarity object c) Get the summary of the model and interpret it carefully d) Get the bi-plot of the model and interpret it carefully

Do the following in R Studio with R script to knit PDF output: a) Create a dataset with 200 random cases and five random variables with your roll number as random seed b) Fit a hierarchical clustering model using single linkage and get the dendogram for this model c) Fit a hierarchical clustering model using complete linkage and get the dendogram for this model d) Fit a hierarchical clustering model using average linkage and get the dendogram for this model e) Find the best hierarchical clustering model for this data and locate the number of clusters for it

OR

Use the four variables of "USArrests" data file into R Studio and do as follows with R script to knit PDF output: a) Fit a k-means clustering model in the data with k=2, plot it in the single graph and interpret it carefully b) Add cluster centers for the plot of clusters formed with k=2 and interpret it carefully c) Fit a k-means clustering model in the data with k=3, plot it in the single graph and interpret it carefully d) Add cluster centers for the plot of clusters formed with k=3 and interpret it carefully