machine learning.md

machine learning

machine learning is looking for a function .It can do some maths and spite an answer .

approaches:

1. supervised learning
2. unsupervised learning
3. reinforce learning

regression

simple learner model : it is a method to predict the variable y based on independent variable x.

multiple learner model : it is a met to predict the variable y based on multiple independent variable .

step

1. model

we suppose a linear model

y=b+∑wixi

w and b are parameter

2. loss function

we can use loss function to evaluate our model . loss function is a measure of how good a prediction model dose in term of being able to predict the expected outcome .

3. gradient decent

gradient decent is a iterative optimization algorithm for finding a local minimum of a differentiable function .

classification

logistics regression : logistics regression is used for classification problems. It can predict the group to which the object belongs to. It give you a binary outcome between 0 and 1.

K nearest neighbours : KNN is used for classification and regression . Theory is that similar things are near to each other .