regression analysis jared dean as quoted in big data, data mining, and machine learning – from my...

17
Regression Analysis • Jared Dean as quoted in Big Data, Data Mining, and Machine Learning – From my experience, regression is the most dominant force in driving business decisions today. Regression analysis has many useful characteristics; one is the easy interpretation of results. Regression concepts are widely understood, and the methodology is well developed such that a well-tuned regression model by a skilled practitioner can outperform many algorithms that are gaining popularity from the machine learning discipline.

Upload: scot-waters

Post on 25-Dec-2015

221 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Regression Analysis Jared Dean as quoted in Big Data, Data Mining, and Machine Learning – From my experience, regression is the most dominant force in

Regression Analysis

• Jared Dean as quoted in Big Data, Data Mining, and Machine Learning– From my experience, regression is the most dominant

force in driving business decisions today. Regression analysis has many useful characteristics; one is the easy interpretation of results. Regression concepts are widely understood, and the methodology is well developed such that a well-tuned regression model by a skilled practitioner can outperform many algorithms that are gaining popularity from the machine learning discipline.

Page 2: Regression Analysis Jared Dean as quoted in Big Data, Data Mining, and Machine Learning – From my experience, regression is the most dominant force in

Simple Linear Regression

Page 4: Regression Analysis Jared Dean as quoted in Big Data, Data Mining, and Machine Learning – From my experience, regression is the most dominant force in

Simple Linear Regression

Linear regression calculator

Compute the equation for the least-squares, best-fit line through the 5 points {(2,2), (0,0), (-2,-2), (-1,1), (1,-1)}

Page 5: Regression Analysis Jared Dean as quoted in Big Data, Data Mining, and Machine Learning – From my experience, regression is the most dominant force in

General Regression – not just straight line Polynomial Curve Fitting

Bishop Textbook – Chapter 1

Page 6: Regression Analysis Jared Dean as quoted in Big Data, Data Mining, and Machine Learning – From my experience, regression is the most dominant force in

Polynomial Curve FittingBishop Textbook – Chapter 1

Page 7: Regression Analysis Jared Dean as quoted in Big Data, Data Mining, and Machine Learning – From my experience, regression is the most dominant force in

Polynomial Curve FittingBishop Textbook – Chapter 1

Page 8: Regression Analysis Jared Dean as quoted in Big Data, Data Mining, and Machine Learning – From my experience, regression is the most dominant force in

General Regression – not just straight lineLeast Squares Approximation - Algorithms Textbook

Page 9: Regression Analysis Jared Dean as quoted in Big Data, Data Mining, and Machine Learning – From my experience, regression is the most dominant force in
Page 10: Regression Analysis Jared Dean as quoted in Big Data, Data Mining, and Machine Learning – From my experience, regression is the most dominant force in
Page 11: Regression Analysis Jared Dean as quoted in Big Data, Data Mining, and Machine Learning – From my experience, regression is the most dominant force in
Page 12: Regression Analysis Jared Dean as quoted in Big Data, Data Mining, and Machine Learning – From my experience, regression is the most dominant force in
Page 13: Regression Analysis Jared Dean as quoted in Big Data, Data Mining, and Machine Learning – From my experience, regression is the most dominant force in
Page 14: Regression Analysis Jared Dean as quoted in Big Data, Data Mining, and Machine Learning – From my experience, regression is the most dominant force in
Page 15: Regression Analysis Jared Dean as quoted in Big Data, Data Mining, and Machine Learning – From my experience, regression is the most dominant force in

Least Squares Approximation

• Pseudoinverse of Matrix A - Linear Example• On Whiteboard

Compute the equation for the least-squares, best-fit line through the 5 points {(2,2), (0,0), (-2,-2), (-1,1), (1,-1)}

Page 16: Regression Analysis Jared Dean as quoted in Big Data, Data Mining, and Machine Learning – From my experience, regression is the most dominant force in

Linear Regression versusPrincipal Component Analysis

Linear Regression Principal Component Analysis

Reference

Page 17: Regression Analysis Jared Dean as quoted in Big Data, Data Mining, and Machine Learning – From my experience, regression is the most dominant force in

Linear Regression versusReverse Linear Regression

Linear Regression Reverse Linear Regression

Reference