Each programmer must know the top ten machine learning algorithm

As a program developer era Ai artificial intelligence, whether or not to participate in the big data and artificial intelligence technology related to development work, we need to understand some commonly used machine learning algorithms, today the candle education teacher must give you recommend 10 algorithm.

1, linear regression: In the field of machine learning and statistics, linear regression is the most widely used algorithm is the most easy to get started.
2, Logistic Regression: Logistic regression is a technique to draw from the field of statistics, mainly deal with binary classification.
3, linear discriminant analysis: linear discriminant analysis algorithm (LDA) is used to deal with two or more classes of preferred classification technique.
4, Classification and Regression Trees: Decision Trees are an important class of predictive modeling algorithms, it's learning fast speed make predictions soon.
5, Naive Bayes: Naive Bayes is a simple and the effect is very good predictive modeling algorithm, it is assumed that each input variable independent of each other, this algorithm when dealing with a large number of complex issues is very good.
6, K Nearest Neighbor algorithm: K Nearest Neighbor (KNN) algorithm is very simple, KNN model that represents the entire training data set, sometimes, KNN or require a lot of memory space to store all the data, but only if you need to perform calculations predict .
7, learning vector quantization: KNN with different learning vector quantization algorithm (of LVQ) to select the desired data training examples, these examples and accurately learn, no need to process the entire training data set.
8, SVM: Support Vector Machine (SVM) is one of the areas of machine learning can be directly used in the most powerful classifier.
9, bagged and random forest method: An integrated random forest machine learning algorithms in machine learning is widely used.
10, Boosting and AdaBoost: Boosting is an attempt to exploit a large number of integrated technology of weak classifiers; AdaBoost is the first algorithm for real Boosting the development of binary classification.

Above the 10 machine-learning algorithms as the industry commonly used algorithm, has a very important position, is recommended every programmer should be simple or in-depth understanding, it is best to apply it in real business scenarios.

Guess you like

Origin blog.51cto.com/14355900/2401708