Handling imbalanced text data in NLP - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to import the class used for oversampling minority classes.

NLP

from imblearn.over_sampling import [1]

Drag options to blanks, or click blank then click option'

ATfidfVectorizer

BRandomOverSampler

CSMOTE

DCountVectorizer

Attempts:

3 left

💡 Hint

Common Mistakes

Confusing vectorizers with oversampling methods.

Using RandomOverSampler instead of SMOTE.

The SMOTE class is commonly used to oversample minority classes in imbalanced datasets.

2fill in blank

medium

Complete the code to convert text data into numerical features using TF-IDF.

NLP

from sklearn.feature_extraction.text import [1]
tfidf = [1](stop_words='english')
X = tfidf.fit_transform(texts)

Drag options to blanks, or click blank then click option'

ACountVectorizer

BTfidfVectorizer

CSMOTE

DRandomOverSampler

Attempts:

3 left

💡 Hint

Common Mistakes

Using CountVectorizer which only counts words.

Trying to use oversampling classes here.

TfidfVectorizer converts text into numerical features by calculating term frequency-inverse document frequency.

3fill in blank

hard

Fix the error in applying SMOTE to the feature matrix X and labels y.

NLP

smote = SMOTE(random_state=42)
X_resampled, y_resampled = smote.fit_resample([1], y)

Drag options to blanks, or click blank then click option'

AX

By

Csmote

Drandom_state

Attempts:

3 left

💡 Hint

Common Mistakes

Swapping the order of X and y.

Passing the SMOTE object instead of data.

The fit_resample method takes the feature matrix X and labels y. Passing y as the first argument is incorrect.

4fill in blank

hard

Fill both blanks to create a balanced dataset using RandomOverSampler and convert text to features.

NLP

from imblearn.over_sampling import [1]
from sklearn.feature_extraction.text import [2]
ros = [1](random_state=0)
tfidf = [2](stop_words='english')
X = tfidf.fit_transform(texts)
X_resampled, y_resampled = ros.fit_resample(X, y)

Drag options to blanks, or click blank then click option'

ARandomOverSampler

BSMOTE

CTfidfVectorizer

DCountVectorizer

Attempts:

3 left

💡 Hint

Common Mistakes

Mixing SMOTE with RandomOverSampler.

Using CountVectorizer instead of TfidfVectorizer.

RandomOverSampler is used to oversample minority classes. TfidfVectorizer converts text to numerical features.

5fill in blank

hard

Fill all three blanks to create a pipeline that balances data and trains a logistic regression model.

NLP

from imblearn.pipeline import Pipeline
from sklearn.linear_model import [1]
from imblearn.over_sampling import [2]
from sklearn.feature_extraction.text import [3]
pipeline = Pipeline([
    ('vectorizer', [3](stop_words='english')),
    ('oversample', [2](random_state=42)),
    ('classifier', [1]())
])

Drag options to blanks, or click blank then click option'

ALogisticRegression

BRandomOverSampler

CTfidfVectorizer

DSMOTE

Attempts:

3 left

💡 Hint

Common Mistakes

Using SMOTE instead of RandomOverSampler in this pipeline.

Confusing CountVectorizer with TfidfVectorizer.

This pipeline vectorizes text with TfidfVectorizer, balances classes with RandomOverSampler, and trains a LogisticRegression model.