| 
  • If you are citizen of an European Union member nation, you may not use this service unless you are at least 16 years old.

  • Stop wasting time looking for files and revisions. Connect your Gmail, DriveDropbox, and Slack accounts and in less than 2 minutes, Dokkio will automatically organize all your file attachments. Learn more and claim your free account.

View
 

FrontPage

Page history last edited by hoffman.tricia@gmail.com 8 years, 10 months ago

Machine Learning 101

 

Instructors: Dr. Michael Bowles & Dr. Patricia Hoffman

 

Sign Up for the Class at Eventbrite:

http://machinelearning101.eventbrite.com/

 

Overview of the course

Machine Learning 101, deals primarily with supervised learning problems.  Machine Learning 102 will cover unsupervised learning and fault detection. 

 

Both 101 and 102 begin at the level of elementary probability and statistics and from that background survey a broad array of machine learning techniques.  The classes will give participants a working knowledge of these techniques and will leave them prepared to apply those techniques to real problems.  To get the most out of the class, participants will need to work through the homework assignments. 

 

Prerequisites

This class assumes a moderate level of computer programming proficiency.  We will use R (the open source statistics language) for the homework and for the examples in class.  We will cover some of the basics of R and do not assume any prior knowledge of R.  You can find references to how to use R on this website and we will give out sample code during classes that will help get you started. 

 

You'll need some general beginner-level background in probability, calculus, linear algebra and vector calculus.  We will cover most of what is required during the lectures.  The appendices in the back of the Tan text are more than sufficient level for this class. 

 

Machine Learning 101 and 102 can be taken in any any order.  The prerequisites for the two classes are the same.  The second five week session (Machine Learning 102) will culminate in the students giving presentations on papers they have read.

 

Why use R?

We're going to use R as our lingua franca for looking at homework problems, discussing them and comparing different solution approaches.    Load R onto your laptop or desk computer before you come to the first class.   http://cran.r-project.org/  We will include some descriptive material on using R in the first two lectures in order to get everyone up to speed on it. To integrate R with Eclipse click here. References for R are here: References for R Comment on these references here:  Reference for R Comments  More R references

 

General Sequence of Classes:

Machine Learning 101:   Supervised learning

Machine Learning 102Unsupervised Learning and Fault Detection

Text: "Introduction to Data Mining", by Pang-Ning Tan, Michael Steinbach and Vipin Kumar

 

Machine Learning 201:    Advanced Regression Techniques, Generalized Linear Models, and Generalized Additive Models    

Machine Learning 202:   Collaborative Filtering, Bayesian Belief Networks, and Advanced Trees

Text:  "The Elements of Statistical Learning - Data Mining, Inference, and Prediction"  by Trevor Hastie, Robert Tibshirani, and Jerome Friedman

 

Machine Learning Big Data:  Adaptation and execution of machine learning algorithms in the map reduce framework.

 

Machine Learning Text Processing:  Machine learning applied to natural language text documents using statistical algorithms including  indexing, automatic classification (e.g. spam filtering) part of speech identification, topic and modeling, sentiment extraction..

 

Future Topics 

     Data Mining Social Networks

     Text Mining

     Recommender Methods

     Big Data

 

 

Machine Learning 101 Syllabus:   

Week  Topics  Homework  Links 
       
1st Week  Exploring Data    FirstWeekNotes  
    Data Quality     
  Aggregation, Sampling     
   Beginning with R  held at Hacker Dojo 10 AM - Noon Notes
 
 
 
   
2nd Week   Supervised Classification and Prediction     
     General Background  HW #1 Due  SecondWeekNotes 
   Performance Evaluation         
   Trees        
    HW02.pdf    
3rd Week  Regression     
 
Ordinary Least Squares  HW #2 Due   ThirdWeekNotes 
  Ridge Regression  HW03.pdf 
 
       
4th Week  Classification and Regression Techniques     
   

 
k Nearest Neighbors  HW #3 Due  FourthWeekNotes 
  Na├»ve Bayes  HW04 
 
     

 

 

5th Week  Ensemble Methods &  More
   
    Ensemble Techniques
HW #4 Due 

  FifthWeekNotes

  SVM Linear & Nonlinear
 Separable & Nonseparable 
   
       

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

There are more Machine Learning References on Patricia's web site http://patriciahoffmanphd.com/

 

Anyone can read this web site, however only the instructors have permission to edit the site. 

 

Comments (0)

You don't have permission to comment on this page.