NOTE: CHANGED 17 FEB FOR EASE OF USE
Goals
- Predict House prices (Sale_Price) using linear Regression from a modified Ames Housing Dataset
- Prediction accuracy should be above a threshold (to be announced next? week)
Important Links
Submission: 3 parts
- Part 1:
- PDF explaining what you tried, and why you are convinced your solution is the best
- It should include all models you submitted to leaderboard
- Part 2:
- Code You used for cleaning the data and fitting the model
- Part 3:
- predicted.csv
- This part will have (if I can figure it out) a leaderboard enabled based on RMSE
- I will calculate the RMSE, you should only return the predictions
- You may submit more than once
- If you submit a second, third, etc time, you should change your pdf to include what you changed and why
- I have not noted the overall threshold you need to beat. I have added a Submission (FullTest) that you should definitely be able to beat.
Example
# Always good to use
library(tidyverse)
# Load Training Data
house_train <- read.csv("house_train.csv")
# Create your Model
testingModel <- lm(Sale_Price ~ ., data = house_train)
# Predict the test set
house_test <- read.csv("house_test_student.csv") #readcsv
#you may need to clean here depending on what you did above
predicted<-data.frame(unname(predict(testingModel,house_test))) #predict
colnames(predicted)<-"Sale_Price" #rename output col
# Save the csv
write_csv(predicted,"predicted.csv")
You would upload:
1) PDF explanation of what you did, and any overall cleaning/etc you did
2) The script above (which does any data cleaning and training)
3) predicted.csv