1. Statistical Learning: Pattern Classification, Prediction, and Control Peter Bartlett August 2002, UC Berkeley CIS

2. Statistical Learning Pattern classification Key issue: balancing complexity Large margin classifiers Multiclass classification Prediction Control

3. Pattern Classification Given training data, (X 1,Y 1 ),…,(X n,Y n ), find a prediction rule f to minimize For example: –Recognizing objects in an image. –Detecting cancer from a blood serum mass spectrogram.

4. The key issues: Approximation error Estimation error Computation Mass Spectrometer Serum samples Classifier Normal Cancer Pattern Classification

5. Pattern Classification Techniques Linearly parameterized (perceptron) Nonparametric (nearest neighbor) Nonlinearly parameterized –Decision trees –Neural networks Large margin classifiers –Kernel methods –Boosting

6. Pattern Classification A key problem in pattern classification is balancing complexity. A very complex model class has –good approximation, but –poor statistical properties. It is important to know how the model complexity and sample size affect the performance of a classifier. Motivation: Analysis and design of learning algorithms.

7. Minimax Theory for Pattern Classification Optimize the performance in the worst case over classification problems (probability distributions). The minimax performance of a learning system is characterized by the capacity of its model class (VC-dimension). For many model classes, this is closely related to the number of parameters, d: –Linearly parameterized (=d) –Decision trees, neural networks ( ¼ d log d) –Kernel methods, boosting ( ¼ d= 1 ) (Vapnik and Chervonenkis, et al)

8. Regularization Choose model to minimize (empirical error) + (complexity penalty) e.g. complexity approximation error estimation error/penalty

9. Data-Dependent Complexity Estimates More refined: use data to measure (and penalize) complexity. Performance can be significantly better than minimax.

10. Data-Dependent Complexity Estimates Minimizing the regularized criterion is hard. Large margin classifiers solve a simpler version of this optimization problem. For example, –Support vector machines –Adaboost

11. Large Margin Classifiers Two class classification: Y 2 { § 1}. Aim to choose a real-valued function f to minimize risk, Consider the margin, Yf(X): –Positive if sign of f is correct, –Magnitude indicates confidence of prediction.

12. Large Margin Classifiers Choose a convex margin cost function,  Choose f to minimize  -risk,

13. Large Margin Classifiers Adaboost:  (  )=exp(-  ). Support vector machine:  (  )=max(0,1-  ). Other kernel methods:  (  )=max(0,1-  ) 2. Neural networks:  (  )=(1-  ) 2. Logistic regression:  (  )=log(1+exp(-2  )).

14. Large Margin Classifiers: Results Optimizing (convex)  risk: –Computation becomes tractable, –Statistical properties are improved, but –Worse approximation properties. Universal consistency. (Steinwart) Optimal estimation rates; with low noise, better than minimax.

15. More Complex Decision Problems Two-class classification… There are many challenges in more complex decision problems: –Data analysis: multiclass pattern classification, anomaly detection, ranking, clustering, etc. –Prediction –Control

16. Multiclass Pattern Classification In many pattern classification problems, the number of classes is large, and different mistakes have different costs. For example, –Computer vision: Recognizing objects in an image. –Bioinformatics: Predicting gene function from expression profiles.

17. Multiclass Pattern Classification The most successful approach in practice is to convert multiclass classification problems to binary classification problems. Issues: –Code design. –Simultaneously optimize code and classifiers. (Minimal statistical penalty for choosing the code after seeing the data.) –Optimization? More direct approach?

18. Prediction with Structured Data These problems arise, for example, in –Natural language processing, –WWW data analysis, –Bioinformatics (gene/protein networks), and –Analysis of other spatiotemporal signals. Simple heuristics (n-grams, windows) are limited. Challenge: automatically extract from the data relevant structure that is useful for prediction.

19. Control: Sequential Decision Problems Examples: –Robotics. –Choosing the right medical treatment. –In drug discovery, choosing a suitable sequence of candidate drugs to investigate. Approximation/Estimation/Computation: –Complexity of a model class? –Can it be measured from data and experimentation?

20. Control: Sequential Decision Problems Reinforcement learning + control theory: –Adaptive control schemes with performance, stability guarantees? For control problems with discrete action spaces, are there analogs of large margin classifiers, with similar advantages - improving estimation properties by sacrificing approximation properties?

21. Statistical Learning We understand something about two-class pattern classification. More complex decision problems: –prediction –control