Machine Learning Week 1

Machine Learning

Week 1 

What' s meaning for Machine Learning

"A computer program is said to learn from experience E with respect to some class of tasks T and performance measure P, if its performance at tasks in T, as measured by P, improves with experience E."
Experience E = the Experience to do something
Task T= the task to reach the Goal
Measure P= the performance or probability to complete the task
Basically, the machine learning problem are considered to one of two types separately Supervised Learning or Unsupervised Learning

Supervised Learning

Generally, the Supervised Learning problems are categorised into Regression problem and Classification Problem.
In Regression problem: the learning algorithms are aim to predict the result in a continuous outputs.
In Classification problem: the goal for Learning Algorithms is trying to group the input data into discrete category.

Unsupervised Learning 

in unsupervised Learning algorithms, we are able to approach the result without know what the result look likes. In other words, the algorithms will learning from data and find the relationship between variables and produce a result.

Model Representation  

Supervised Learning 

As shown from above picture that, in Supervised Learning Algorithms, we are using the training set data to generate to our Learning algorithms and input the necessary information to produce the predict result.
we are using hypotheses(h(x)) to describe our learning algorithms.

Cost Functions

how to measure the accuracy of our hypotheses, we can use cost function to check.

Cost Functions
J(θ0,θ1)=12m∑i=1m(y^i−yi)2=12m∑i=1m(hθ(xi)−yi)2
The ideal for cost functions is finding the minimal number when we choosing the
θ0,θ1