Euclidean Manhattan and Minkowski Distance || Lesson 57 || Machine Learning || Learning Monkey ||
36.3 هزار بار بازدید -
4 سال پیش
-
#machinelearning
#machinelearning#learningmonkey
In this class, we discuss Euclidean, Manhattan, and Minkowski distance.
We use these Euclidean, Manhattan, and Minkowski Distance in K nearest neighbors.
We use these concepts in many of the distance measure models which we discuss later.
We measure euclidean distance by the square root of the sum of the square of the difference of coordinate values.
E = square root of (x1-x2)^2+(y1-y2)^2.
This measure is a displacement between two points in the coordinate space.
The same concept can be extended to any dimension.
In K nearest neighbors we use this distance measure to find the nearest points.
Manhattan distance:
This is a measure of modulus of difference in coordinates.
M = |x1-x2|+|y1-y2|.
This is a measure of the distance of x coordinate sum distance of y coordinate.
As the number of dimensions increases the measure of euclidean is computationally costly.
In those situations, manhattan distance is good to use in K Nearest neighbors.
The same way Minkowski distance is the generalization of euclidean and manhattan distance.
Given as (sum(|xi-yi|^p))^1/p.
If we place p =2 the above equation turns to euclidean distance.
If we place p=1 the above equation turns to Manhattan distance.
we can place any p-value. but mostly we use euclidean and manhattan distance.
In K nearest neighbors they have given Minkowski formulae. with p-value.
In K neighbor classifier they provide the p-value.
If taken p as 1 manhattan distance.
If taken p as 2 euclidean distance.
Link for playlists:
@learningmonkey
Link for our website: https://learningmonkey.in
Follow us on Facebook @ Facebook: learningmonkey
Follow us on Instagram @ Instagram: learningmonkey1
Follow us on Twitter @ Twitter: _learningmonkey
Mail us @ [email protected]
In this class, we discuss Euclidean, Manhattan, and Minkowski distance.
We use these Euclidean, Manhattan, and Minkowski Distance in K nearest neighbors.
We use these concepts in many of the distance measure models which we discuss later.
We measure euclidean distance by the square root of the sum of the square of the difference of coordinate values.
E = square root of (x1-x2)^2+(y1-y2)^2.
This measure is a displacement between two points in the coordinate space.
The same concept can be extended to any dimension.
In K nearest neighbors we use this distance measure to find the nearest points.
Manhattan distance:
This is a measure of modulus of difference in coordinates.
M = |x1-x2|+|y1-y2|.
This is a measure of the distance of x coordinate sum distance of y coordinate.
As the number of dimensions increases the measure of euclidean is computationally costly.
In those situations, manhattan distance is good to use in K Nearest neighbors.
The same way Minkowski distance is the generalization of euclidean and manhattan distance.
Given as (sum(|xi-yi|^p))^1/p.
If we place p =2 the above equation turns to euclidean distance.
If we place p=1 the above equation turns to Manhattan distance.
we can place any p-value. but mostly we use euclidean and manhattan distance.
In K nearest neighbors they have given Minkowski formulae. with p-value.
In K neighbor classifier they provide the p-value.
If taken p as 1 manhattan distance.
If taken p as 2 euclidean distance.
Link for playlists:
@learningmonkey
Link for our website: https://learningmonkey.in
Follow us on Facebook @ Facebook: learningmonkey
Follow us on Instagram @ Instagram: learningmonkey1
Follow us on Twitter @ Twitter: _learningmonkey
Mail us @ [email protected]
4 سال پیش
در تاریخ 1399/02/25 منتشر شده
است.
36,321
بـار بازدید شده