吴恩达·Machine Learning || chap10 Advice for applying machine learning简记

10 Advice for applying machine learning

10-1 Deciding what to try next

Debugging a learning algorithm

Suppose you have implemented regularized linear regression to predict housing prices.

$$J(\theta )=\frac{1}{2m}[\sum _{i=1}^m(h_0(x^{(i)})-y^{(i)}{)}^2+\lambda \sum _{j=1}^m{\theta }_j^2]$$

However, when you test your hypothesis on a new set of houses, you find that it makes unacceptably large errors in its predictions. What should you try next?

• Get more training examples

• Try smaller sets if features

• Try getting additional features

• Try adding polynomial features

• Try decreasing $\lambda$

• Try increasing $\lambda$

Machine learning diagnostic:

Diagnostic: A test that you can run to gain insight what is/isn’ t working with a learning algorithm, and gain guidance as to how best to improve its performance.

Diagnostics can take time to implement, but doing so can be a very good use of your time.

10-2 Evaluating a hypothesis

Evaluating your hypothesis

Fails to generalize to new examples not in training set

Training/testing procedure for linear regression

• Learn parameter $\theta$ from training data(minimizing training error ($J(\theta )$))
• Compute test set error :$J_{test}(\theta )$

classification problem:

• Learn parameter $\theta$ from training data
• Compute test set error:

​ $$J_{test}(\theta )=-\frac{1}{m_{test}}\sum _{i=1}^{m_{test}}{y}_{test}^{(i)}\log h_{\theta }(x_{test}^{(i)})+(1-y_{test}{)}^{(i)}log{h}_{\theta }(x_{test}^{(i)})$$

• Miscalssification error (0/1 miscalssification error):

10-3 Model selection and training/validation/test sets

Overfitting example

Once parameters ${\theta }_0,{\theta }_1,\cdots ,{\theta }_n$ were fit to some set of data (training set), the error of the parameters as measured on that data(the training error $J(\theta )$ is likely to be lower than the actual generalization error.

Model selection

parameter d=degree of polynomial

How well does the model generalize? Report test set error $J_{test}({\theta }^{(d)})$

Problem: $J_{test}({\theta }^{(d)})$ is likely to be an optimistic estimate of generalization error l.e. our extra parameter(d = degree of polynomial)is fit to test set.

Evaluating your hypothesis

Training set 60%

cross validation 20%

test set 20%

Training/validation/test error

Training error:

$$J_{train}(\theta )=\frac{1}{2m}\sum _{i=1}^m(h_{\theta }(x^{(i)})-y^{(i)}{)}^2$$

Cross validation error:

$$J_{cv}(\theta )=\frac{1}{2m_{cv}}\sum _{i=1}^{m_v}(h_{\theta }(x_{cv}^{(i)})-y_{cv}^{(i)}{)}^2$$

Test error:

$$J_{test}(\theta )=\frac{1}{2m_{test}}\sum _{i=1}^{m_{test}}(h_{\theta }(x_{test}^{(i)})-y_{test}^{(i)}{)}^2$$

Estimate generalization error for test set $J_{test}({\theta }^{(4)})$

10-4 Diagnosing bias vs. variance

Bias/variance 偏差/方差

Training error:

Crossing validation error:

Diagnosing bias vs variance
Suppose your learning algorithm is performing less well than you were hoping. $(J_{cv}(\theta )or{J}_{test}(\theta )ishigh)$ Is it a bias problem or a variance problem?

Bias (underfit):$J_{train}(\theta )$ will be high;$J_{cv}(\theta )\approx J_{train}(\theta )$

Variance(Overfit):$J_{train}(\theta )$ will be low;$J_{cv}(\theta )>>J_{train}(\theta )$

10-5 Regularization and bias/variance

Linear regression with regularization

Model:

$$h_{\theta }(x)={\theta }_0+{\theta }_{1}x+{\theta }_2{x}^2+{\theta }_3{x}^3+{\theta }_4{x}^4$$

$$J(\theta )=\frac{1}{2m}[\sum _{i=1}^m(h_{\theta }(x^{(i)})-y^{(i)}{)}^2+\lambda \sum _{j=1}^n{\theta }_j^2]$$

High bias——Large $\lambda$

Just right——intermediate $\lambda$

High variance——small $\lambda$

Choosing the regularization parameter $\lambda$

​ the smallest $J_{cv}({\theta }^{(i)})$ error

Bias/variance as a function of the regularization parameter $\lambda$

10-6 Learning curves

Learning curves

$J_{train}(\theta )$

$J_{cv}(\theta )$

High bias: more and more close

If a learning algorithm is suffering from high bias, getting more training data will not (by itself) help much

High variance: large gap

If a learning algorithm is suffering from high variance, getting more training data is likely to help

10-7 Deciding what to try next (revisited)

Debugging a learning algorithm:

• Get more training examples $⟶$fixed high variance
• Try smaller sets if features$⟶$fixed high variance
• Try getting additional features$⟶$fixed high bias
• Try adding polynomial features$⟶$$⟶$fixed high bias
• Try decreasing $\lambda$$⟶$fixed high bias
• Try increasing $\lambda$$⟶$fixed high variance

Neural networks and overfitting