Bringing it all together

In addition to the distance-based learning anomaly detection pipeline you created in the last exercise, you want to also support a feature-based learning one with one-class SVM. You decide to extract two features: first, the length of the string, and, second, a numerical encoding of the first letter of the string, obtained using the function LabelEncoder() described in Chapter 1. To ensure a fair comparison, you will input the outlier scores into an AUC calculation. The following have been imported: LabelEncoder(), roc_auc_score() as auc() and OneClassSVM. The data is available as a pandas data frame called proteins with two columns, label and seq, and two classes, IMMUNE SYSTEM and VIRUS. A fitted LoF detector is available as lof_detector.

This exercise is part of the course

Designing Machine Learning Workflows in Python

View Course

Exercise instructions

For a string s, len(s) returns its length. Apply that to the seq column to obtain a new column len.
For a string s, list(s) returns a list of its characters. Use this to extract the first letter of each sequence, and encode it using LabelEncoder().
LoF scores are in the negative_outlier_factor_ attribute. Compute their AUC.
Fit a 1-class SVM to a data frame with only len and first as columns. Extract the scores and assess both the LoF scores and the SVM scores using AUC.

Hands-on interactive exercise

Have a go at this exercise by completing this sample code.

# Create a feature that contains the length of the string
proteins['len'] = proteins['seq'].apply(____)

# Create a feature encoding the first letter of the string
proteins['first'] =  ____.____(
  proteins['seq'].apply(____))

# Extract scores from the fitted LoF object, compute its AUC
scores_lof = lof_detector.____
print(____(proteins['label']==____, scores_lof))

# Fit a 1-class SVM, extract its scores, and compute its AUC
svm = ____.____(proteins[['len', 'first']])
scores_svm = svm.____(proteins[['len', 'first']])
print(____(proteins['label']==____, scores_svm))

Edit and Run Code

Designing Machine Learning Workflows in Python

AdvancedSkill Level

4.8+

55 reviews

In the previous chapters you established a solid foundation in supervised learning, complete with knowledge of deploying models in production but always assumed you a labeled dataset would be available for your analysis. In this chapter, you take on the challenge of modeling data without any, or with very few, labels. This takes you into a journey into anomaly detection, a kind of unsupervised modeling, as well as distance-based learning, where beliefs about what constitutes similarity between two examples can be used in place of labels to help you achieve levels of accuracy comparable to a supervised workflow. Upon completing this chapter, you will clearly stand out from the crowd of data scientists in confidently knowing what tools to use to modify your workflow in order to overcome common real-world challenges.

Exercise 1: Anomaly detection Exercise 2: A simple outlier Exercise 3: LoF contamination Exercise 4: Novelty detection Exercise 5: A simple novelty Exercise 6: Three novelty detectors Exercise 7: Contamination revisited Exercise 8: Distance-based learning Exercise 9: Find the neighbor Exercise 10: Not all metrics agree Exercise 11: Unstructured data Exercise 12: Restricted Levenshtein Exercise 13: Bringing it all together

Current Exercise

Exercise 14: Concluding remarks