WEKA 3.1.8 - PowerPoint PPT Presentation

About This Presentation
Title:

WEKA 3.1.8

Description:

Another classifier is one of the parameters. Boosting, bagging, logitboost, metacost, stacking, etc. Methods: SMO, k-NN, LWR, logistic... – PowerPoint PPT presentation

Number of Views:182
Avg rating:3.0/5.0
Slides: 30
Provided by: AleksJ5
Category:
Tags: weka | bagging

less

Transcript and Presenter's Notes

Title: WEKA 3.1.8


1
WEKA 3.1.8
Waikato Environment for Knowledge
Analysis http//www.cs.waikato.ac.nz/ml/weka/
2
Program
  • A quick overview of WEKA functionality
  • WEKA demonstration
  • Visualization and EG2000

3
WEKA
  • Command line interface
  • Explorer
  • Filters
  • Learning (clustering, classification, assoc.
    Rules)
  • Attribute selection
  • Basic visualization
  • Experimenter
  • Testing a particular method on a set of data
  • Various criteria for comparison of results

4
Classification
  • Meta-methods
  • Another classifier is one of the parameters.
  • Boosting, bagging, logitboost, metacost,
    stacking, etc.
  • Methods
  • SMO, k-NN, LWR, logistic...
  • C4.5, m5...
  • Associative rules.

5
Filters
  • Adding/removing attributes
  • Attribute value substitution
  • Discretization (MDL, Kononenko, etc.)
  • Time series filters (delta, shift)
  • Sampling, randomization
  • Missing value management
  • Normalization and other numeric transformations

6
Attribute Selection
  • Search methods
  • Best-first.
  • Genetic.
  • Ranking...
  • Measures
  • ReliefF.
  • information gain.
  • Gain ratio

7
Other Features
  • Probability estimators
  • Method evaluators
  • Margin curve (difference between the predicted
    class probability and max probability of other
    classes)
  • Threshold curve (varying class proportions)
  • Statistics for binary class domains
  • Precision correct p/total predicted P
  • F-measure 2TPprecision/(TPprecision)

8
My Humble Opinion
  • Good
  • Simplicity
  • Features
  • GUI
  • Bad
  • Kitchen sink syndrome
  • Strange object hierarchy
  • Incomplete documentation

9
(No Transcript)
10
(No Transcript)
11
(No Transcript)
12
Eurographics
  • Kake uporabne ideje?

13
Vizualizacija
  • Pristop
  • Za nek tip podatkov
  • najdemo dober nacin prikaza.
  • Pozor
  • Dobra vizualizacija namesto pravil ali
    odlocitvenih dreves.
  • Dober vmesnik namesto avtomatske analize
    podatkov.
  • K sreci
  • Za vizualizacijo predvsem uporabljajo
    vecdimenzionalne prostorske podatke.

14
Deleži
Mozaicni diagram
15
Deleži 2
16
Raztreseni diagram
  • tudi planetni ali mehurcni diagram
  • William Playfair (1759-1823)
  • The Commercial and Political Atlas, 1786, uvedel
    tudi
  • crtni diagram (line)
  • stolpcni diagram (bar)
  • tortni diagram (pie)

17
Interaktivnost
18
Interaktivnost 2
19
Dimenzije
  • figura (glyph)
  • 5 dimenzij
  • prostor
  • barva
  • velikost

20
Priprava vizualizacije
dataflow
21
Osnovni elementi
  • Obseg
  • Pozicija
  • Velikost
  • Sivina
  • Razlika
  • Smer
  • Barva
  • Oblika
  • Vzorec

Neurejeni Urejeni Kolicinski
22
Iskanje povezanosti
23
(No Transcript)
24
(No Transcript)
25
(No Transcript)
26
(No Transcript)
27
(No Transcript)
28
(No Transcript)
29
(No Transcript)
Write a Comment
User Comments (0)
About PowerShow.com