Lasso, Support Vector Machines, Generalized linear models Kenneth D. Harris 20/5/15.

Slides:



Advertisements
Similar presentations
Regularized risk minimization
Advertisements

1 1 Chapter 5: Multiple Regression 5.1 Fitting a Multiple Regression Model 5.2 Fitting a Multiple Regression Model with Interactions 5.3 Generating and.
Linear Regression.
Regularization David Kauchak CS 451 – Fall 2013.
Prediction with Regression
Biointelligence Laboratory, Seoul National University
Pattern Recognition and Machine Learning
Computer vision: models, learning and inference Chapter 8 Regression.
R OBERTO B ATTITI, M AURO B RUNATO. The LION Way: Machine Learning plus Intelligent Optimization. LIONlab, University of Trento, Italy, Feb 2014.
Chapter 2: Lasso for linear models
Multivariate Methods Pattern Recognition and Hypothesis Testing.
Coefficient Path Algorithms Karl Sjöstrand Informatics and Mathematical Modelling, DTU.
Lecture 14 – Neural Networks
The Kernel Trick Kenneth D. Harris 3/6/15.
Ridge regression and Bayesian linear regression Kenneth D. Harris 6/5/15.
Linear Methods for Regression Dept. Computer Science & Engineering, Shanghai Jiao Tong University.
Predictions 1. Multiple linear regression Kenneth D. Harris April 29, 2015.
Machine Learning CUNY Graduate Center Lecture 3: Linear Regression.
Lecture outline Support vector machines. Support Vector Machines Find a linear hyperplane (decision boundary) that will separate the data.
Measures of Association Deepak Khazanchi Chapter 18.
Review Rong Jin. Comparison of Different Classification Models  The goal of all classifiers Predicating class label y for an input x Estimate p(y|x)
Classification and Prediction: Regression Analysis
PATTERN RECOGNITION AND MACHINE LEARNING
Machine Learning CUNY Graduate Center Lecture 3: Linear Regression.
2.4: Cautions about Regression and Correlation. Cautions: Regression & Correlation Correlation measures only linear association. Extrapolation often produces.
1 Logistic Regression Adapted from: Tom Mitchell’s Machine Learning Book Evan Wei Xiang and Qiang Yang.
Non-Linear Models. Non-Linear Growth models many models cannot be transformed into a linear model The Mechanistic Growth Model Equation: or (ignoring.
Mathematical formulation XIAO LIYING. Mathematical formulation.
Jeff Howbert Introduction to Machine Learning Winter Regression Linear Regression.
Session 10. Applied Regression -- Prof. Juran2 Outline Binary Logistic Regression Why? –Theoretical and practical difficulties in using regular (continuous)
LOGISTIC REGRESSION David Kauchak CS451 – Fall 2013.
Logistic Regression Database Marketing Instructor: N. Kumar.
CS Statistical Machine learning Lecture 18 Yuan (Alan) Qi Purdue CS Oct
Overview of the final test for CSC Overview PART A: 7 easy questions –You should answer 5 of them. If you answer more we will select 5 at random.
Jeff Howbert Introduction to Machine Learning Winter Regression Linear Regression Regression Trees.
Sparse Kernel Methods 1 Sparse Kernel Methods for Classification and Regression October 17, 2007 Kyungchul Park SKKU.
M Machine Learning F# and Accord.net.
CORRECTIONS L2 regularization ||w|| 2 2, not ||w|| 2 Show second derivative is positive or negative on exams, or show convex – Latter is easier (e.g. x.
CpSc 881: Machine Learning
A P STATISTICS LESSON 3 – 3 (DAY 3) A P STATISTICS LESSON 3 – 3 (DAY 3) RISIDUALS.
From OLS to Generalized Regression Chong Ho Yu (I am regressing)
Logistic Regression & Elastic Net
Machine Learning CUNY Graduate Center Lecture 6: Linear Regression II.
Section 3.2 Linear Models: Building Linear Functions from Data.
Chapter 14 Introduction to Regression Analysis. Objectives Regression Analysis Uses of Regression Analysis Method of Least Squares Difference between.
Support Vector Regression in Marketing Georgi Nalbantov.
Data Mining: Concepts and Techniques1 Prediction Prediction vs. classification Classification predicts categorical class label Prediction predicts continuous-valued.
Outlier Detection Identifying anomalous values in the real- world database is important both for improving the quality of original data and for reducing.
Machine Learning & Deep Learning
Generalized regression techniques
CSE 4705 Artificial Intelligence
Linear Regression (continued)
Regularized risk minimization
Roberto Battiti, Mauro Brunato
Multiple Regression A curvilinear relationship between one variable and the values of two or more other independent variables. Y = intercept + (slope1.
Simple Linear Regression
Least-Squares Regression
What is Regression Analysis?
The Bias Variance Tradeoff and Regularization
Biointelligence Laboratory, Seoul National University
Contact: Machine Learning – (Linear) Regression Wilson Mckerrow (Fenyo lab postdoc) Contact:
Least-Squares Regression
Support Vector Machine I
y = mx + b Linear Regression line of best fit REMEMBER:
Basis Expansions and Generalized Additive Models (1)
EQUATION 4.1 Relationship Between One Dependent and One Independent Variable: Simple Regression Analysis.
Machine learning overview
4.2 Cautions about Correlation and Regression
A medical researcher wishes to determine how the dosage (in mg) of a drug affects the heart rate of the patient. Find the correlation coefficient & interpret.
Cases. Simple Regression Linear Multiple Regression.
Presentation transcript:

Lasso, Support Vector Machines, Generalized linear models Kenneth D. Harris 20/5/15

Multiple linear regression What are you predicting? Data typeContinuous Dimensionality1 What are you predicting it from? Data typeContinuous Dimensionalityp How many data points do you have?Enough What sort of prediction do you need?Single best guess What sort of relationship can you assume?Linear

Ridge regression What are you predicting? Data typeContinuous Dimensionality1 What are you predicting it from? Data typeContinuous Dimensionalityp How many data points do you have?Not enough What sort of prediction do you need?Single best guess What sort of relationship can you assume?Linear

Regression as a probability model What are you predicting? Data typeContinuous Dimensionality1 What are you predicting it from? Data typeContinuous Dimensionalityp How many data points do you have?Not enough What sort of prediction do you need?Probability distribution What sort of relationship can you assume?Linear

Different data types What are you predicting? Data typeDiscrete, integer, whatever Dimensionality1 What are you predicting it from? Data typeContinuous Dimensionalityp How many data points do you have?Not enough What sort of prediction do you need?Single best guess What sort of relationship can you assume?Linear – nonlinear

Ridge regression Fit quality Penalty

“Regularization path” for ridge regression

Changing the penalty

The LASSO Fit quality Penalty

LASSO regularization path Most weights are exactly zero “sparse solution”, selects a small number of explanatory variables This can help avoid overfitting when p>>N Models are easier to interpret – but remember there is no proof of causation. Path is piecewise-linear

Elastic net

Predicting other types of data Fit quality Penalty

Support vector machine f E

Errors vs. margins

Generalized linear models What are you predicting? Data typeDiscrete, integer, whatever Dimensionality1 What are you predicting it from? Data typeContinuous Dimensionalityp How many data points do you have?Not enough What sort of prediction do you need?Probability distribution What sort of relationship can you assume?Linear – nonlinear

Generalized linear models

Example: logistic regression f P(y; f)

Logistic regression loss function

Poisson regression

What to read; what software to use