Overfitting in Tree Induction

Report
Chapter 5
Overfitting and Its Avoidance
指 導 教 授 : 徐 立 群
學
生
教 授
: R16014101 陳怡齊
R16011234 吳年鑫
Overfitting
 即「過適」、「超適」或稱「過度擬合」
 意指在調適一個model時,使用過多參數。對比於可取得的資料總量
來說,一個荒謬的model只要足夠複雜,是可以完美地適應(fit)資料。
 不合乎一般化 (Generalization)
 違反奧卡姆剃刀(Occam’s Razor ) 原則
Overfitting & Generalization
 A extreme example –
 Customer churn or non-churn
 Training data & Holdout data
Overfitting Examined
• Holdout Data and Fitting Graphs A fitting graph shows the accuracy of a model as a function of
complexity .
Figure 1. A typical fitting graph.
Overfitting Examined
 Base rate  What would b be ?
Figure 2. A fitting graph for the customer churn (table) model.
Overfitting in Tree Induction
 Decision tree induction
 overfitting starts to
 the “sweet spot” in the graph .
Figure 3. A typical fitting graph for tree induction.
Overfitting in Mathematical Functions
 We add more Xi, the function becomes more and more complicated.
 Each Xi has a corresponding Wi, which is a learned parameter of the
model .
 Two dimensions you can fit a line to any two points and in three
dimensions you can fit a plane to any three points .
 This concept generalizes: as you increase the dimensionality, you can
perfectly fit larger and larger sets of arbitrary points .
Example: Overfitting Linear Functions
Data:sepal width, petal width
Types:Iris Setosa, Iris Versicolor
Two different separation lines:
a. Logistic regression
b. Support vector machine
Figure 4
Example: Overfitting Linear Functions
Figure 4
Figure 5
Example: Overfitting Linear Functions
Figure 6
Figure 7
From Holdout Evaluation to Cross-Validation
Holdout Evaluation Splits the data
into only one training and one
holdout set.
Cross-validation
computes
its
estimates over all the data by
performing multiple splits and
systematically swapping out samples
for testing. ( k folds, typically k would
be 5 or 10. )
The Churn Dataset Revisited
“Example: Addressing the Churn Problem with Tree Induction” in Chapter 3.
 The logistic regression models
show slightly lower average
accuracy (64.1%) and with
higher variation ( standard
deviation of 1.3 )
 Average accuracy of the folds
with classification trees is
68.6%—significantly lower than
our previous measurement of
73%. ( the standard deviation of
the fold accuracies is 1.1 )
 Classification trees may be
preferable
to
logistic
regression because of their
greater
stability
and
performance.
Learning Curves
 The generalization performance of data-driven
modeling generally improves as more training data
become available.
Overfitting Avoidance & Complexity Control
Concept in Tree Induction :
 Tree induction commonly uses two techniques to avoid overfitting. These
strategies are :
 (i) to stop growing the tree before it gets too complex, and
 (ii) to grow the tree until it is too large, then “prune” it back, reducing its size (and
thereby its complexity).
Methods in Tree Induction :
 To limit tree size is to specify a minimum number of instances that must be
present in a leaf.
 Hypothesis test ( P-value )
Overfitting Avoidance & Complexity Control
General Method for Avoiding Overfitting
 Compare the best model we can build from one family (say, classification
trees) against the best model from another family (say, logistic regression).
Nested holdout testing
 Select the best model by assess by
having a complexity of 122 nodes ( the
sweet spot).
Training subset
Training set
 Induce a new tree with 122 nodes
from the whole, original training data.
Validation set
Test set
( hold out )
Final hold out
Overfitting Avoidance & Complexity Control
Nested Cross-Validation
Sequential Forward Selection
Training set
Test set
Original data

similar documents