# What type of algorithm is naive Bayes?

The naive Bayes method is a supervised learning technique that uses Bayes’s theorem to solve classification issues. It is mostly used for text classification, including a high-dimensional training set.

The Naive-Bayes Classifier is one of the simplest and most successful classification algorithms, aiding in the development of fast Machine Learning models capable of making rapid predictions. It is a classification algorithm, meaning it makes predictions based on the likelihood of an item.

Popular applications of the naive Bayes algorithm include spam filtering, sentimental analysis, and article classification.

## Neutral Gaussian Bayes

This form of the naive Bayes algorithm is used when data is continuous. All variables are assumed to have a normality test. If you have variables that lack this quality, you may want to turn them into features with normal distribution.

## Multinomial Naive Bayes

This is used when the frequency is represented by the characteristics.

Assume you have a TXT file where you recover all the specific words and build numerous features, each of which represents the word count in the document. Frequency is a feature in that situation. Here, you use Multinomial Naive Bayes.

It disregards the absence of the characteristics, so if the frequency of a feature is zero, its probability of occurrence is also zero. Multinomial Naive Bayes disregards that feature. It is effective for text categorization issues.

## Bernoulli Naive Bayes

This is used for binary characteristics. Instead of utilizing the periodicity of the word, discrete 1s and 0s that signify the existence or absence of a characteristic is used. In this scenario, the features will be binary, and the Bernoulli Naive Bayes will be used.

This technique also penalizes the absence of a characteristic.

Benefits of naive Bayes:

• Easy to implement and comprehend.
• Substantially quicker than other algorithms since it just calculates probability.