Selecting Data For Modeling

Two Approaches

• Dot notation

  • Select the "prediction target"

• Selecting with a column list

  • Selects the "features"

Dot Notation (choosing pred. target)

• Select column we want to predict (Prediction Target)

# By convention, this is named y
y = model.Column

Choosing "features"

• Features are columns inputted into model

features = [price, name, clothes]
# By convention, this is named X
X = model[features]

Steps to building and using a model

• Define

  • What type of model? (eg. decision tree)

• Fit

  • Capture patterns from model

• Predict

• Evaluate

  • Determine accuracy of model's predictions

Example using scikit-learn

If you do not specify random_state number, model may allow for some randomness in model training

from sklearn.tree import DecisionTreeRegressor

# Define model. Specify a number for random_state to ensure same results each run
model = DecisionTreeRegressor(random_state=1)

# Fit model
model.fit(X, y)

>>DecisionTreeRegressor(criterion='mse', max_depth=None, max_features=None,
           max_leaf_nodes=None, min_impurity_decrease=0.0,
           min_impurity_split=None, min_samples_leaf=1,
           min_samples_split=2, min_weight_fraction_leaf=0.0,
           presort=False, random_state=1, splitter='best')

print("Making predictions for the following 5 houses:")
print(X.head())
print("The predictions are")
print(melbourne_model.predict(X.head()))

>>> Making predictions for the following 5 houses:
   Rooms  Bathroom  Landsize  Lattitude  Longtitude
1      2       1.0     156.0   -37.8079    144.9934
2      3       2.0     134.0   -37.8093    144.9944
4      4       1.0     120.0   -37.8072    144.9941
6      3       2.0     245.0   -37.8024    144.9993
7      2       1.0     256.0   -37.8060    144.9954
The predictions are
[1035000. 1465000. 1600000. 1876000. 1636000.]