King County House Price Prediction

Background

• King County is located in the U.S. state of Washington. The population was 2,252,782 in the 2019 census estimate, making it the most populous county in Washington, and the 12th-most populous in the United States. The county seat is Seattle, also the state's most populous city.
• How much reasonable price for house in King County?

Goal

Predict a reasonable house price in King County

About Data

We got this data from Kaggle: https://www.kaggle.com/harlfoxem/housesalesprediction

Feature Description

• id : Unique id for each home sold
• date : Date of the home sale
• price : Price of each home sold
• bedrooms : Number of bedrooms
• bathrooms : Number of bathrooms, where .5 accounts for a room with a toilet but no shower
• sqft_living : Square footage of the apartments interior living space
• sqft_lot : Square footage of land space
• floors : Number of floors
• waterfront : A dummy variable for whether the apartment was overlooking the waterfront or not
• view : An index from 0 to 4 of how good the view of the property was
• condition : An index from 1 to 5 on the condition of the apartment
• grade : An index from 1 to 13, where 1-3 falls short of building construction and design, 7 has an average level of construction and design, and 11-13 have a high quality level of construction and design
• sqft_above : The square footage of the interior housing space that is above ground level
• sqft_basement : The square footage of the interior housing space that is below ground level
• yr_built : The year the house was initally built
• yr_renovated : The year of the house's last renovation
• zipcode : What zipcode area the house is in
• lat : Latitude
• long : Longitude
• sqft_living15 : The square footage of interior housing living space for the nearest 15 neighbors
• sqft_lot15 : The square footage of the land lots of the nearest 15 neighbors

Table of Content

• Data Exploration
• Data Analysis and Data Visualization
  • Univariate
  • Bivariate
• EDA Conclusion
• EDA Recommendation
• Data Preparation
• Base Model
• Feature Engineering 1
• Feature Engineering 2
• Hyperparameter Tuning
• ML Conclusion

EDA Conclusion

• Houses with price above median mostly are located in the center and nearby of city
• Houses with price under median mostly are located far from city
• Houses with 0 - 3 bedrooms mostly under median price
• Houses with more than 3 bedrooms mostly above median price
• Houses with 0 - 2 bathrooms mostly under median price
• Houses with more than 2 bathrooms mostly above median price
• Middle-low (225 - 644 sqft) and Middle-up (645 - 1291 sqft) houses mostly under median price
• Luxury (> 1291 sqft) houses mostly above median price
• Houses with 1 floor mostly under median price
• Houses with more than 1 floor mostly above median price
• Houses without waterfront almost balance between number of under and above median price
• Houses with waterfront mostly above median price
• Houses without view mostly under median price
• Houses with except very bad view mostly above median price
• Houses with very bad, bad, and very good condition mostly under median price
• Houses with best condition mostly above median price
• Houses with good condition are almost balance between under and above median price
• Houses with bad quality and average quality mostly under median price
• Houses with high quality mostly above median price
• Houses without basement mostly under median price
• Houses with basement mostly above median price
• House with basement mostly luxury house (>1292 sqft)
• Houses that renovated mostly above median price
• Houses that not renovated mostly under median price
• Number between young and old house almost balance
• Young houses mostly above median price
• Old houses mostly under median price

EDA Recommendation

• Houses that are located in the center and nearby city can priced above median
• Houses that are located far from city can priced under median
• Houses with 0 - 3 bedrooms can priced under median price
• Houses with more than 3 bedrooms can priced above median price
• Houses with 0-2 bathrooms can priced under median price
• Houses with more than 2 bathrooms can proced above median
• Middle-low and Middle-up houses can priced under median
• Luxury houses can priced above median
• Houses with 1 floor can priced under median price
• Houses with more than 1 floor can priced above median price
• Houses without waterfront can priced under or above median price depends on other variables
• Houses with waterfront can priced above median price
• Houses without view can priced under median price
• Houses with view can priced above median price
• Houses with very bad, bad, and very good condition can priced under median price
• Houses with good and best condition can priced above median price
• Houses with bad and average quality can priced under median price
• Houses with high quality can priced above median price
• Houses without basement can priced under median price
• Houses with basement can priced above median price
• Houses that not renovated can priced under median price
• Houses that renovated can priced above median price Young (<= 40 years old) houses can priced above median price Old (more than 40 years old) houses can priced under median price

ML Conclusion

• Ridge:
  • Best score from ridge is still Ridge + Polynomial + price feature transformation, fit in 0.82
• KNN:
  • Best score from KNN is KNN + Hyperparameter Tuning, test score reach 0.85 although the score are overfitting among train and test
• XGBoost:
  • Best score from XGB is XGBoost + Hyperparameter Tuning, test score reach 0.91 although the score are overfitting among train and test. Also this is highest test score than other algorithms