So what is naive Bayesian algorithm? In fact, naive Bayes belongs to generative model, that is, generative model and discriminant model, which mainly depends on whether joint distribution is needed. This algorithm is relatively simple, you just need to do a bunch of counting. If you pay attention to the hypothesis of conditional independence, the convergence speed of naive Bayesian classifier will be faster than that of discriminant models such as logistic regression, so you only need less training data. Even if the assumption of NB conditional independence is not established, NB classifier still performs well in practice. Its main disadvantage is that it can't learn the interaction between features. In terms of R in mRMR, it is feature redundancy.
So what are the advantages of naive Bayesian algorithm? This algorithm has five advantages. Firstly, Naive Bayesian model originated from classical mathematical theory, which has a solid mathematical foundation and stable classification efficiency. The second is to have a high speed when training and querying a lot. Even if a very large-scale training set is used, each item usually has relatively few feature numbers, and the training and classification of items are only mathematical operations of feature probability. Thirdly, it performs well on small-scale data, can handle multi-classification tasks, and is suitable for incremental training (that is, it can train new samples in real time). Fourthly, it is insensitive to missing data, and the algorithm is relatively simple, which is often used for text classification. Fifth, Naive Bayes is easy to understand the explanation of the results.
Of course, the shortcomings of naive Bayesian algorithm are also obvious. Naive Bayesian algorithm has four shortcomings. The first is to calculate the prior probability. The second is the error rate of classification decision. The third is very sensitive to the expression of input data. The fourth is to assume that the sample attributes are independent, so if the sample attributes are related, the effect is not good.
So what is the application field of naive Bayes? In fact, Naive Bayesian algorithm is widely used in fraud detection. Of course, we can also use naive Bayes algorithm to judge whether an email is spam. Naive Bayesian algorithm can also be used to judge the category of articles, and Bayesian algorithm can also be used to judge whether a paragraph expresses positive emotions or negative emotions. It can be seen that naive Bayesian algorithm is a very practical algorithm.
In this paper, we introduced some knowledge about the advantages and disadvantages of naive Bayesian algorithm. I believe that through the explanation of this knowledge, you have a certain understanding of naive Bayesian algorithm, and I hope this article can help you.