Slide 1 Additive Models and Trees

Lecture Notes for CMPUT 466/551

Nilanjan Ray

Principal Source: Department of Statistics, CMU

Slide 2 ### Topics to cover

- GAM: Generalized Additive Models
- CART: Classification and Regression Trees
- MARS: Multiple Adaptive Regression Splines

Slide 3 ### Generalized Additive Models

What is GAM?

The functions fjare smoothing functions in general, such as splines, kernel

functions, linear functions, and so on…

Each function could be different, e.g., f1 can be linear, f2 can be a natural

spline, etc.

Compare GAM with Linear Basis Expansions (Ch. 5 of [HTF])

Similarities? Dissimilarities?

Any similarity (in principle) with Naïve Bayes model?

Slide 4 ### Smoothing Functions in GAM

- Non-parametric functions (linear smoother)
- Smoothing splines (Basis expansion)
- Simple k-nearest neighbor (raw moving average)
- Locally weighted average by using kernel weighting
- Local linear regression, local polynomial regression

- Linear functions
- Functions of more than one variables (interaction term)
- Example:

Slide 5 ### Learning GAM: Backfitting

Backfitting algorithm

Slide 6 ### Backfitting: Points to Ponder

Computational Advantage?

Convergence?

How to choose fitting functions?

Slide 7 ### Example: Generalized Logistic Regression

Model:

Slide 8 Additive Logistic Regression: Backfitting

Fitting logistic regression (P99)

Fitting additive logistic regression (P262)

1. where

1.

2.

2.

Iterate:

Iterate:

a.

a.

b.

b.

c.

Using weighted least squares to fit a linear model to zi with weights wi, give new estimates

c. Using weighted backfitting algorithm to fit an additive model to zi with weights wi, give new estimates

3. Continue step 2 until converge

3.Continue step 2 until converge

Slide 9 ### SPAM Detection via Additive Logistic Regression

- Input variables (predictors):
- 48 quantitative variables: percentage of words in the email that match a given word. Examples include business, address, internet, etc.
- 6 quantitative variables: percentage of characters in the email that match a given character, such as ‘ch;’, ch(, etc.
- The average length of uninterrupted sequences of capital letters
- The length of the longest uninterrupted sequence of capital letters
- The sum of length of uninterrupted length of capital letters

- Output variable: SPAM (1) or Email (0)
- fj’s are taken as cubic smoothing splines

Slide 10 ### SPAM Detection: Results

Sensitivity: Probability of predicting spam given true state is spam =

Specificity: Probability of predicting email given true state is email =

Slide 11 ### GAM: Summary

- Useful flexible extensions of linear models
- Backfitting algorithm is simple and modular
- Interpretability of the predictors (input variables) are not obscured
- Not suitable for very large data mining applications (why?)

Slide 12 ### CART

- Overview
- Principle behind: Divide and conquer
- Partition the feature space into a set of rectangles
- For simplicity, use recursive binary partition

- Fit a simple model (e.g. constant) for each rectangle
- Classification and Regression Trees (CART)
- Regress Trees
- Classification Trees

- Popular in medical applications

Slide 13 ### CART

- An example (in regression case):

Slide 14 ### Basic Issues in Tree-based Methods

- How to grow a tree?
- How large should we grow the tree?

Slide 15 ### Regression Trees

- Partition the space into M regions: R1, R2, …, RM.

Note that this is still an additive model

Slide 16 ### Regression Trees– Grow the Tree

- The best partition: to minimize the sum of squared error:
- Finding the global minimum is computationally infeasible
- Greedy algorithm: at each level choose variable j and value s as:
- The greedy algorithm makes the tree unstable
- The error made at the upper level will be propagated to the lower level

Slide 17 ### Regression Tree – how large should we grow the tree ?

- Trade-off between bias and variance
- Very large tree: overfit (low bias, high variance)
- Small tree (low variance, high bias): might not capture the structure

- Strategies:
- 1: split only when we can decrease the error (usually short-sighted)
- 2: Cost-complexity pruning (preferred)

Slide 18 Penalty on the complexity/size of the tree

Cost: sum of squared errors

### Regression Tree - Pruning

- Cost-complexity pruning:
- Pruning: collapsing some internal nodes
- Cost complexity:
- Choose best alpha: weakest link pruning (p.270, [HTF])
- Each time collapse an internal node which add smallest error
- Choose from this tree sequence the best one by cross-validation

Slide 19 ### Classification Trees

- Classify the observations in node m to the major class in the node:
- Pmk is the proportion of observation of class k in node m

- Define impurity for a node:
- Misclassification error:
- Entropy:
- Gini index :

Slide 20 ### Classification Trees

- Entropy and Gini are more sensitive
- To grow the tree: use Entropy or Gini
- To prune the tree: use Misclassification rate (or any other method)

Node impurity measures versus class proportion for 2-class problem

Slide 21 ### Tree-based Methods: Discussions

- Categorical Predictors
- Problem: Consider splits of sub tree t into tL and tR based on categorical predictor x which has q possible values: 2(q-1)-1 ways !
- Treat the categorical predictor as ordered by say proportion of class 1

Slide 22 ### Tree-based Methods: Discussions

- Linear Combination Splits
- Split the node based on
- Improve the predictive power
- Hurt interpretability

- Instability of Trees
- Inherited from the hierarchical nature
- Bagging (section 8.7 of [HTF]) can reduce the variance

Slide 23 ### Bootstrap Trees

Construct B number of trees from B bootstrap samples– bootstrap trees

Slide 24 ### Bootstrap Trees

Slide 25 ### Bagging The Bootstrap Trees

is computed from the bth bootstrap sample

in this case a tree

Bagging reduces the variance of the original tree by aggregation

Slide 26 ### Bagged Tree Performance

Majority vote

Average

Slide 27 ### MARS

- In multi-dimensional spline the basis functions grow exponentially– curse of dimensionality
- A partial remedy is a greedy forward search algorithm
- Create a simple basis-construction dictionary
- Construct basis functions on-the-fly
- Choose the best-fit basis function at each step

Slide 28 ### Basis functions

- 1-dim linear spline (t represents the knot)
- Basis collections C:
|C| = 2 * N * p

Slide 29 ### The MARS procedure (1st stage)

- Initialize basis set M with a constant function
- Form candidates (cross-product of M with set C)
- Add the best-fit basis pair (decrease residual error the most) into M
- Repeat from step 2 (until e.g. |M| >= threshold)

M (new)

M (old)

C

Slide 30 ### The MARS procedure (2nd stage)

The final model M typically overfits the data

=>Need to reduce the model size (# of terms)

Backward deletion procedure

Slide 31 ### Generalized Cross Validation (GCV)

- M(.) measures effective # of parameters:
- r: # of linearly independent basis functions
- K: # of knots selected
- c = 3

Slide 32 ### Discussion

- Piecewise linear reflected basis
- Allow operation on local region
- Fitting N reflected basis pairs takes O(N) instead of O(N^2)
- Left-part is zero, right-part differs by a constant

X[i-1]

X[i]

X[i+1]

X[i+2]

Slide 33 Slide 34 ### MARS & CART relationship

IF