(20 pts)
Submit your results to Lab 1 by responding here. Your response should include:
numeric_variables = ['GrLivArea']
nominal_variables = ['Neighborhood']
qual_variables = ['KitchenQual']
housing_predictions_demo.csv (19.8 KB)
My Score:
0.196455
numeric_variables = ['GrLivArea', 'OverallQual', 'YearBuilt', 'OverallCond', 'GarageCars', 'Fireplaces', 'FullBath', 'BsmtFullBath', 'TotalBsmtSF', 'YearRemodAdd', 'TotRmsAbvGrd']
nominal_variables = ['Neighborhood']
qual_variables = ['KitchenQual']
housing_predictions_demo.csv (19.8 KB)
Score: 0.14427434825801208
numeric_variables = ['GrLivArea', 'OverallQual','YearBuilt','OverallCond','GarageCars','BsmtFullBath','Fireplaces',
'TotalBsmtSF', 'YearRemodAdd', 'TotRmsAbvGrd', 'FullBath', 'ScreenPorch', 'LotArea', 'WoodDeckSF', 'BsmtFinSF1']
nominal_variables = ['Neighborhood']
qual_variables = ['KitchenQual']
target_variables = ['SalePrice']
housing_predictions_demo.csv (19.8 KB)
numeric_variables = ['GrLivArea', 'YearBuilt', 'GarageCars', 'OverallQual', 'OverallCond']
nominal_variables = ['Neighborhood']
qual_variables = ['KitchenQual']
target_variables = ['SalePrice']
numeric_variables = ['GrLivArea', 'OverallQual', 'YearBuilt', 'OverallCond', 'GarageCars']
nominal_variables = ['Neighborhood']
qual_variables = ['KitchenQual']
target_variables = ['SalePrice']
housing_predictions_demo (1).csv (19.8 KB)
numeric_variables = ['GrLivArea', 'MSSubClass','GarageCars', 'OverallQual','OverallCond']
nominal_variables = ['Neighborhood','YearBuilt']
qual_variables = ['KitchenQual']
target_variables = ['SalePrice']
housing_predictions_demo.csv (19.8 KB)
Score: 0.15877368818470677
numeric_variables = ['OverallQual','GrLivArea','YearBuilt','OverallCond','GarageCars','BsmtFullBath']
nominal_variables = ['Neighborhood']
qual_variables = ['KitchenQual']
My score was 0.1499
housing_predictions_demo (15).csv (19.8 KB)
numeric_variables = ['GrLivArea','OverallQual','GarageCars','Fireplaces','ScreenPorch','3SsnPorch','WoodDeckSF','EnclosedPorch','OverallCond','YearRemodAdd','LotArea']
nominal_variables = ['Neighborhood']
qual_variables = ['KitchenQual']
housing_predictions_demo (13).csv (19.8 KB)
numeric_variables = ['GrLivArea','YearBuilt', 'OverallQual', 'OverallCond', 'GarageCars']
nominal_variables = ['Neighborhood']
qual_variables = ['KitchenQual']
target_variables = ['SalePrice']
housing_predictions_demo (1).csv (19.8 KB)
Score: 0.15417995829350553
numeric_variables = [ 'GrLivArea', 'OverallQual', 'YearBuilt', 'OverallCond',
'GarageCars', 'BsmtFullBath', 'Fireplaces',
'TotalBsmtSF', 'YearRemodAdd', 'TotRmsAbvGrd',
'FullBath', 'ScreenPorch', 'LotArea', 'WoodDeckSF',
'BsmtFinSF1']
nominal_variables = ['Neighborhood', 'BldgType', 'RoofMatl']
qual_variables = ['KitchenQual']
housing_predictions_demo.csv (19.8 KB)
My Score:
0.13724343241678663
Variables
numeric_variables = ['GrLivArea','OverallQual','NHmedian','YearBuilt','OverallCond','GarageCars','BsmtFullBath','Fireplaces', 'TotRmsAbvGrd','ScreenPorch','TotalBsmtSF','YearRemodAdd','FullBath','LotArea']
nominal_variables = ['Neighborhood']
qual_variables = ['KitchenQual']
target_variables = ['SalePrice']
The NHmedian variable is the median sale price calculated for each neighborhood--location, location, location.
Score
0.14289
housing_predictions_demo.csv (19.8 KB)
numeric_variables = ['GrLivArea', 'OverallQual', 'YearBuilt', 'OverallCond', 'GarageArea', 'BsmtFullBath', 'Fireplaces', 'TotalBsmtSF',
'YearRemodAdd', 'TotRmsAbvGrd', 'FullBath', 'ScreenPorch', 'LotArea']
nominal_variables = ['Neighborhood']
qual_variables = ['KitchenQual']
target_variables = ['SalePrice']
housing_predictions_demo.csv (19.8 KB)
numeric_variables = ['GrLivArea', 'YearBuilt', 'OverallCond', 'Fireplaces', 'FullBath' ]
nominal_variables = ['Neighborhood']
qual_variables = ['KitchenQual']
target_variables = ['SalePrice']
our score is 0.1690946121350628.
housing_predictions_demo (3).csv (19.8 KB)
numeric_variables = ['GrLivArea','OverallQual','YearBuilt','GarageCars','BsmtFullBath',
'Fireplaces','TotalBsmtSF','YearRemodAdd','TotRmsAbvGrd','FullBath','ScreenPorch','LotArea']
housing_predictions_demo.csv (19.8 KB)
0.14423910325793002
My changes were
numeric_variables = ['OverallQual','OverallCond','GarageCars','LotArea','TotalBsmtSF','GarageArea','WoodDeckSF','OpenPorchSF','PoolArea']
nominal_variables = ['Neighborhood','RoofMatl','Utilities','Foundation','BsmtCond']
qual_variables = ['KitchenQual']
housing_predictions_demo.csv (19.8 KB)
numeric_variables = ['GrLivArea', 'OverallQual', 'YearBuilt', 'OverallCond', 'GarageCars', 'BsmtFullBath', 'Fireplaces', 'TotalBsmtSF', 'YearRemodAdd', 'TotRmsAbvGrd', 'FullBath', 'ScreenPorch', 'LotArea', 'WoodDeckSF', 'BsmtFinSF1', 'HalfBath', '1stFlrSF', 'EnclosedPorch', '3SsnPorch', 'GarageArea', 'BsmtHalfBath', 'BsmtFinSF2', 'MasVnrArea', 'LowQualFinSF', 'BsmtUnfSF', '2ndFlrSF', 'MoSold', 'BedroomAbvGr', 'OpenPorchSF', 'MiscVal', 'YrSold', 'KitchenAbvFr', 'PoolArea', 'MSSubClass']
nominal_variables = ['Neighborhood']
qual_variables = ['KitchenQual']
target_variables = ['SalePrice']
Copy of HousingPriceRegressionLab2026.ipynb - Colab
Mean Squared Error: 0.13224937762230213
@User 028
What's your score???
Sorry, forgot to add it. Mine was 0.16678281096727288
The easiest way to start improving the score was by adding more variables to consider in our regression. I started with numeric which made a massive improvement but limited me to a score roughly of 0.1475. By adding nominal variables I was able to reduce it to roughly 0.1425
numeric_variables = [ 'GrLivArea', 'OverallQual', 'YearBuilt', 'OverallCond',
'GarageCars', 'BsmtFullBath', 'Fireplaces',
'TotalBsmtSF', 'YearRemodAdd', 'TotRmsAbvGrd',
'FullBath', 'ScreenPorch', 'LotArea', 'WoodDeckSF',
'BsmtFinSF1', 'MSSubClass', 'PoolArea', 'YrSold', 'MiscVal',
]
nominal_variables = ['Neighborhood', 'BldgType', 'RoofMatl', 'HouseStyle', 'Foundation', 'SaleCondition',]
The next big jump I was able to do was by adding more quality variables. The initial code would not allow for this due to the shape of the array so I had to map the quality order to each index in the array and encoding them to quality order. This got me down to 0.1356.
qual_variables = ['KitchenQual', 'ExterQual', 'BsmtQual', 'HeatingQC', 'ExterCond', 'BsmtCond']
qual_encoder = OrdinalEncoder(categories=[quality_order] * len(qual_variables))
Finally, due to the increase in variables it did turn out to be useful to use RidgeCV so I uncommented:
regress = RidgeCV(
alphas=np.logspace(-1, 1, 100)
)
and
regress.alpha_
I'm pretty sure there is a better way to narrow down the noise by consolidating similar attributes such as total square footage instead of each sqft attribute individually but I couldn't get that to work properly. There also might be some logspace values that could possibly help in the RidgeCV but that also didn't work out for me
Final score: 0.13166469694390048
Lab01_MML.csv (19.8 KB)
numeric_variables = ['GrLivArea', 'OverallQual', 'YearBuilt', 'BedroomAbvGr', 'OverallCond', 'GarageCars', 'BsmtFullBath', 'Fireplaces', 'TotalBsmtSF', 'YearRemodAdd', 'FullBath', 'LotArea', 'WoodDeckSF', 'HalfBath', 'MasVnrArea']
nominal_variables = ['Neighborhood','MSSubClass', 'Street', 'Condition1', 'Functional', 'CentralAir', 'RoofStyle', 'Foundation', 'BldgType', 'Heating', 'MSZoning', 'HouseStyle', 'LotShape']
qual_variables = ['KitchenQual', 'ExterQual', 'HeatingQC', 'BsmtQual', 'GarageQual', 'BsmtExposure', 'BsmtFinType1', 'BsmtFinType2']
Maggie.csv (19.8 KB)
score is 0.13383794318864528
Definitely diminishing returns on just adding variables.
numeric_variables = ['GrLivArea', 'YearBuilt', 'OverallCond', 'GarageCars', 'Fireplaces', 'BsmtFullBath', 'TotalBsmtSF', 'TotRmsAbvGrd', 'KitchenAbvGr']
nominal_variables = ['Neighborhood']
qual_variables = ['KitchenQual']
target_variables = ['SalePrice']
My score was a 0.15369876296498602
housing_predictions_demo (4).csv (19.8 KB)
numeric_variables = ['GrLivArea', 'OverallQual','YearBuilt','OverallCond','GarageCars','BsmtFullBath','Fireplaces',
'TotalBsmtSF', 'YearRemodAdd', 'TotRmsAbvGrd', 'FullBath', 'ScreenPorch', 'LotArea', 'WoodDeckSF', 'BsmtFinSF1']
nominal_variables = ['Neighborhood', 'BldgType', 'RoofMatl', 'HouseStyle']
qual_variables = ['KitchenQual']
target_variables = ['SalePrice']
housing_predictions_demo (1).csv (19.8 KB)
