AI and ML for Finance Practitioners

Quiz LO 7.2.4

 Test your knowledge of LO 7.2.4

 25%

Question 1 of 4

1. Is the following statement about ‘Binary NB ‘ generally correct?

Statement:
In sentiment classification, “whether a word occurs or not” matters more than its frequency. Therefore, it often improves performance to clip the word counts in each document at 1 as shown in the figure below. If a word is found in two documents then it is counted as 2, etc. This variant is known as binary multinomial naive Bayes or binary NB.

Question 1 of 4

Question 2 of 4

2. Is the following statement about ‘ the reasons why binary NB might improve results relative to the standard Naïve Bayes approach ‘ generally correct?

Statement:
In sentiment classification, whether a word occurs or not matters more than its frequency. This means that Naive Bayes classifiers would perform better in sentiment classification (more computationally efficient) as they capture the presence of a word in a document with the binary indicator 1 for “present” and 0 for “not-present”, thereby avoiding counting all duplicates.

Question 2 of 4

Question 3 of 4

3. According to the method for capturing the sentiment of a negative sentence, how should the sentence “I don’t like this book” be changed?


I) “I don’t Not_like this book”

II) “I Not_don’t like this book”

III) “I don’t Not_like Not_this Not_book”

Question 3 of 4

Question 4 of 4

4. Sentiment lexicons are lists of words that are pre-annotated with positive or negative sentiment. Which of these is not a popular lexicons?

I) General Inquirer.

II) LIWC.

III) BNN.

IV) MPQA Subjectivity Lexicon.

Question 4 of 4


 

x  Powerful Protection for WordPress, from Shield Security
This Site Is Protected By
Shield Security