Machine Learning week 9 quiz: Anomaly Detection

Anomaly Detection

5 试题

1. 

For which of the following problems would anomaly detection be a suitable algorithm?

Given data from credit card transactions, classify each transaction according to type of purchase (for example: food, transportation, clothing).

From a large set of primary care patient records, identify individuals who might have unusual health conditions.

In a computer chip fabrication plant, identify microchips that might be defective.

From a large set of hospital patient records, predict which patients have a particular disease (say, the flu).

2. 

Suppose you have trained an anomaly detection system for fraud detection, and your system that flags anomalies when  p(x)  is less than  ε , and you find on the cross-validation set that it mis-flagging far too many good transactions as fradulent. What should you do?

Decrease  ε

Increase  ε

3. 

Suppose you are developing an anomaly detection system to catch manufacturing defects in airplane engines. You model uses

p(x)=∏nj=1p(xj;μj,σ2j).

You have two features  x1  = vibration intensity, and  x2  = heat generated. Both  x1  and  x2  take on values between 0 and 1 (and are strictly greater than 0), and for most "normal" engines you expect that  x1≈x2 . One of the suspected anomalies is that a flawed engine may vibrate very intensely even without generating much heat (large  x1 , small  x2 ), even though the particular values of  x1 and  x2  may not fall outside their typical ranges of values. What additional feature  x3  should you create to capture these types of anomalies:

x3=x1×x2

x3=x1+x2

x3=x21×x2

x3=x1x2

4. 

Which of the following are true? Check all that apply.

If you have a large labeled training set with many positive examples and many negative examples, the anomaly detection algorithm will likely perform just as well as a supervised learning algorithm such as an SVM.

When choosing features for an anomaly detection system, it is a good idea to look for features that take on unusually large or small values for (mainly the) anomalous examples.

If you are developing an anomaly detection system, there is no way to make use of labeled data to improve your system.

If you do not have any labeled data (or if all your data has label  y=0 ), then is is still possible to learn  p(x) , but it may be harder to evaluate the system or choose a good value of  ϵ .

5. 

You have a 1-D dataset  {x(1),…,x(m)}  and you want to detect outliers in the dataset. You first plot the dataset and it looks like this:

Suppose you fit the gaussian distribution parameters  μ1  and  σ21  to this dataset. Which of the following values for  μ1  and  σ21  might you get?

μ1=−3,σ21=4

μ1=−6,σ21=4

μ1=−3,σ21=2

μ1=−6,σ21=2