Machine Learning Practice Quiz 1: 16 July 2023 (May 2023 term)

Select all that apply.

1. A

   The range of scale of a numerical feature in the data

2. B

   The distribution of a numerical feature in the data

3. C

   The null values present in a feature of the data

4. D

   The correlation between features and labels in the data

Select all that apply.

1. A

   Data has some numerical and some categorical features.

2. B

   Data has only categorical features and all of them are nominal.

3. C

   Data has only categorical features, however, some features are ordinal and some are nominal.

4. D

   Data has only numerical features, and all of them are uniformly distributed in range of 0 and 1 (both inclusive).

1data = pd.read_csv('dataset.csv')
2X = data[data.columns[:-1]]
3y = data[data.columns[-1]]
4X_train, X_test, y_train, y_test = train_test_split(X, y,
5                                                        test_size = 0.8)
6mms = MinMaxScaler()
7X_train['Temperature'] = mms.fit_transform(X_train['Temperature'])
8X_train['Humidity'] = mms.fit_transform(X_train['Humidity'])
9
10
11X_test['Temperature'] = mms.fit_transform(X_test['Temperature'])
12X_test['Humidity'] = mms.fit_transform(X_test['Humidity'])
13
14lr = LinearRegression().fit(X_train,y_train)

Select all that apply.

1. A

   The training set will have 20% of the data, which is a good practice.

2. B

   The training set size is smaller than test set size.

3. C

   The train and test samples are not scaled appropriately.

4. D

   One of the fundamental assumption of Machine Learning, which is, training and test data belong to same distribution, is not upheld.