dontless / machine-learning-foundations-a-case-study-approach Goto Github PK

Can you please run your weak2 assignment[house prices] code again with latest graphlab and also turicreate[below]?[use latest home_data]

import turicreate as tc sales = tc.SFrame('2SnLy-JAEemx8A5HK6Ls8g_0ccbc21b1656423ba7e9d7fe2971879a_home_data.sframe/home_data.sframe') my_features = ['bedrooms', 'bathrooms', 'sqft_living', 'sqft_lot', 'floors', 'zipcode'] advanced_features = ['bedrooms', 'bathrooms', 'sqft_living', 'sqft_lot', 'floors', 'zipcode', 'condition', 'grade', 'waterfront', 'view', 'sqft_above', 'sqft_basement', 'yr_built', 'yr_renovated', 'lat', 'long', 'sqft_living15', 'sqft_lot15'] training_set, test_set = sales.random_split(0.8,seed=0) my_features_model = tc.linear_regression.create(training_set,target='price',features=my_features) TestEvalOutput_my = my_features_model.evaluate(test_set) print(TestEvalOutput_my) adv_features_model = tc.linear_regression.create(training_set,target='price',features=advanced_features) TestEvalOutput_adv = adv_features_model.evaluate(test_set) print(TestEvalOutput_adv) TestEvalOutput_my['rmse'] - TestEvalOutput_adv['rmse']

Because I don't have graphlab and this code gives rmse diff > 25000 and I don't think there's any error in my code. Is it because of upgradation of graphlab to turicreate? But, did they change underlying ML algorithm too?