DAT 537 Assignment 4
title: “Assignment 4”
Copyright By PowCoder代写 加微信 powcoder
output: pdf_document
“`{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
library(cbw)
library(lubridate)
library(ggplot2)
library(dygraphs)
library(xts)
Note: This is an individual assignment. No discussion with a fellow student is allowed. Honor code is in place.
## Question 1
Suppose that you are confronted with a regression problem in which you are unsure about
which covariates (predictors) should be in the model. Suppose that there are 7 covariates
(x1,x2,. . . ,x7) in contention. Your goal is to figure out which of these covariates is/are in the
model. Suppose that you get 500 observations from the population. The X’s are generated
as follows.
Now suppose that the data on the response Y is generated from the following process – you
do not know that this is the process that is generating the data.
1. If you assume that the error is Gaussian, how would you go about finding the best set
of predictors? What is the best model you find under the Gaussian assumption?
2. If you assume that the error is student-t (with possible degrees of freedom equal to
3,4,5, or 6), how would you find the best set of predictors? What is the best model
程序代写 CS代考 加微信: powcoder QQ: 1823890830 Email: powcoder@163.com