Linearly separable and linearly inseparable
Linear separability means that a linear function can be used to separate two types of samples (note that this is a linear function), such as a straight line in two-dimensional space, a plane in three-dimensional space, and a hyperplane in high-dimensional space . The term separable here refers to the separation without a trace of error, and the linear inseparability refers to the phenomenon that some samples will be classified incorrectly when divided by the linear classification surface.
Linear model
The coefficient w before the independent variable x in the multiplication formula. If a w affects only one x, then the model is a linear model, such as y = w 0 + w 1 ∗ xy=w_0 + w_1*xand=w0+w1∗x
when you need to fity = w 0 + w 1 ∗ x + w 2 ∗ x 2 y=w_0+w_1*x+w_2*x^2and=w0+w1∗x+w2∗x2 , you can changex 2 x^2x2 Replace withzzz , that is, using ascending dimension to transform a polynomial regression model into a linear regression model.
Determine whether the data is linearly separable
The convex hull is a convex closed curve (surface) that just encloses all the data.
Check whether the convex hull (convex hull) intersects, which can be used as a basis for judging whether the data is linearly separable.
-
Use quickhull algorithm to find the convex hull of the data
-
The sweepline algorithm determines whether the edges of the convex hull intersect
两个步骤的复杂度都是O(nlogn)
Among them, quickhull has been implemented in the software package qhull (http://www.qhull.org/).