A煤n no tienes acceso a esta clase

Crea una cuenta y contin煤a viendo este curso

Regresi贸n lineal

7/16
Recursos

Aportes 2

Preguntas 0

Ordenar por:

驴Quieres ver m谩s aportes, preguntas y respuestas de la comunidad? Crea una cuenta o inicia sesi贸n.

Escrib铆 un tutorial sobre el modelo de regresi贸n lineal. Puedes encontrar el art铆culo ac谩: https://platzi.com/tutoriales/1766-regresion-python/11159-de-donde-viene-el-algoritmo-de-regresion-lineal/

Linear Regression : parameters

different parameters change what we think of a relation between input features (X) and the output target (y_pred).

y{pred}=w_1x+w_0_

w_1= a change in 鈥渪鈥 means a change in "y"
w_0= the value of 鈥測鈥 if 鈥渪鈥 is 0


Linear Regression : cost function

Mean-square Error

(MSE) Loss :

The cost (E) is the difference between the target value (y_i) and our line predicted value (y_pred)

j\left(w,:w_0\right)=\frac{1}{N}\sum {i=1}^N:\left(y_i-y{i:pred}\right)^2


Linear Regression : actualization rule

We want to minimize the distance (y_i pred) above each point on training data
We change w_0 and w_1 'til find a smaller distance sum is found

_update rule = min J(w_0, w_1) _


pd: i want to add the screenshot, but i don鈥檛 know why doesn鈥檛 let me, it pop up an error: