B06 Naive Bayes

Question 1

What is Naive Bayes in the context of Machine Learning?

Answer 1

Machine learning approach that quantify’s the probability of events and how those probabilities change in the light of additional information.

Question 2

What is probabilistic learning?

Answer 2

Probabilistic learning deals with a set of learning methods which use the observed probabilities of the features within the labeled data to predict the most likely class for unlabeled data

Question 3

The Naïve Bayes algorithm is part of a family of approaches known as :

Answer 3

Probabilistic or bayesian learners.

Question 4

The ________ of an event is the probability that it DID NOT happen.

Answer 4

complement Because events cannot simultaneously happen and not happen, events are mutually exclusive with their complements.

Question 5

Joint Probability

Answer 5

Question 6

Conditional Probability (Bayes Theorem)

For dependent events A and B , the probability of event A,
given that event B occurred is described with a conditional
probability:

Answer 6

Question 7

What is NAIVE Bayes?

Answer 7

Naïve Bayes is a simple application of Bayes’ theorem to
classification, with the “naïve” assumption that all features
in the dataset are equally important and independent
within a class (Class Conditional independence).

In other words, events are assumed to be independent so
long as they are conditioned on the same class value.

Question 8

What is Additive Smoothing?

Answer 8

Given an observation from a multinomial
distribution with trials, additive or Laplace smoothing is
used to smooth the data by adding a pseudocount ( alpha) to
the number of observed and unobserved cases in order to
change the expected probability so as to avoid zerofrequency
problems.

Question 9

Strengths of Naive Bayes?

Answer 9

• Simple, fast and effective.
• Works well with noisy and missing data.
• Useful for both continuous and discrete features.
• Works well with both little or large training data.
• Estimated probabilities are easy to obtain.

Question 10

Weaknesses of Naive Bayes?

Answer 10

• Assumes that features are equally important and independent.
• Not suitable for datasets with many continuous features.
• Correlated features degrade performance.
• Estimated probabilities may seem less reliable than predicted classes

Question 11

Applications of Naive Bayes?

Answer 11

• Junk email filtering (spam).
• Network intrusion or anomaly detection.
• Medical condition diagnosis.