In today’s era, having an education qualification of maths and computer lands you to the most demanding job - machine learning engineer. But there could not be a machine learning engineer without the basic knowledge of the top 10 machine learning algorithms.
Classifying a web page, document, emails manually is laborious work. Even sorting out our mail is tedious work. But had we thought how come spam filtering takes place. Naive Bayes classifier is the main principle behind spam filtering. Fundamental theory behind Naive Bayes classifier is:
P (A)
Clustering is an unsupervised classification algorithm. As the name suggests, Clustering means making clusters of similar objects, i.e. similar objects are paired together. K means clustering helps search engines to cluster web pages by similarity.
Support Vector Machines commonly known as SVM, is a supervised algorithm based on the optimization principle. Increasing the distance between the two hyperplanes is the postulate of this algorithm. SVM can be quite beneficial for an investor as it works very well with comparing stocks.
Linear Regression is the simplest algorithm demonstrating a linear relationship between dependent and independent variables. The mathematical theory behind this algorithm is Linear Algebra. It works great in business development. It proves useful in increasing sales.
Logistic Regression is linear regression restricted to sigmoid function or tanh function. Logistic regression is quite useful while modelling probabilities.
Tree is a structure with leaves and branches. Similar is the decision tree. The topic further extends to Random Forest which is the practical impli- cation of Decision tree. It gives a very good accuracy. It is well suited in the situations when some of the data is missing.
The second most complex algorithm is ANN(Artificial Neural Network). Though complex but it appears to be quite useful algorithm. ANN is quite useful in Human facial recognition.