Machine Learning

What is Machine Learning

Arthur Samuel

• the field of study that gives computers the ability to learn without being explicitly programmed.

  Tom Mitchell

• 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.
• Example: playing checkers.
  E = the experience of playing many games of checkers
  T = the task of playing checkers.
  P = the probability that the program will win the next game.
  In general, any machine learning problem can be assigned to one of two broad classifications:
  Supervised learning and Unsupervised learning.

  Supervised Learning

  Supervised Learning

• In supervised learning, we are given a data set and already know what our correct output should look like, having the idea that there is a relationship between the input and the output.
• Category
  • Classification
    • In a classification problem, we are instead trying to predict results in a discrete output. In other words, we are trying to map input variables into discrete categories.
  • Regression
    • In a regression problem, we are trying to predict results within a continuous output, meaning that we are trying to map input variables to some continuous function.

      Linear Regression

• Gradient Descent For Linear Regression

  • Math Formula

      - 

  • Multiple Features
    • Defination
      • Linear regression with multiple variables is also known as “multivariate linear regression”.

    • Math Formula

      -
    • Feature Scaling
      • How
        • We can speed up gradient descent by having each of our input values in roughly the same range.
      • Why
        • This is because θ will descend quickly on small ranges and slowly on large ranges, and so will oscillate inefficiently down to the optimum when the variables are very uneven.
      • • Si is the range of Xi
    • Learning Rate
      • To summarize:
        If α is too small: slow convergence.
        If α is too large ,may not decrease on every iteration and thus may not converge
      • Tips
        • No matter how small the α is ， it will reach reach minimum in the end。
    • Features and Polynomial Regression
      • You can also use multiple feature to try to analysis tit by using one variable, If so you can fit any shape of function you want

        Classification

• Defination
  • We don’t need to get the number result of each result , What we should do is just to try to get the bool of whether it in this category or not , or where does it belong to?
• How
  • To attempt classification, one method is to use linear regression and map all predictions greater than 0.5 as a 1 and all less than 0.5 as a 0. However, this method doesn’t work well because classification is not actually a linear function.
• Hypothesis Representation

  • Sigmoid Function

    • Usage
      • It can help us avoid the influence of some big data and help us judge the class which it belongs to

  • Math Formula

      - 

  • Result

    • Explaination
      • The formula above is the possible when y equals to 1
• Decision Boundary

  • How to choose y

    • Unsupervised Learning

      Unsupervised Learning

• Unsupervised learning allows us to approach problems with little or no idea what our results should look like. We can derive structure from data where we don’t necessarily know the effect of the variables.

  Model

  x ——- input

  y ——- output
  h ——- hypothesis function

  Cost Function

  Defination

• We can measure the accuracy of our hypothesis function by using a cost function.

  Form

• Squared error function

• Tips
  • We use 1/2m is simplify our compute level
• Result
  • What we should do is to calculate the result which can make the J(x) minmum

    Gradient Descent

    Defination

• The method to estimate the parameters in the hypothesis function.

  How

• We should try to get the minimum θ so we should change it along with the function changing, all the θ must change at once

• • About

  • α is the learning speed
• Tips
  • Use tamp variable to store the number ,and change them at once

    Normal Equation

    Defination

• A algorithm to try to consult out the derivative of the J(x) = 0 and use it as the θ

  Math Furmula

• Difference

赏

mia mia mia