Here are the results of a "Curse of Dimensionality" homework assignment for Terran Lane's Introduction to Machine Learning class. Pretty pictures, interesting results, and a good exercise in explicit parallelism with R.
Exercise 5.2 Improves the Logistic Regression implementation done in Exercise 4 by adding a regularization parameter that reduces the problem of over-fitting. We will be using Newton's Method.
Here's the data we want to fit.
# linear regression # load the data mydata = read.csv("http://spreadsheets.google.com/pub?key=0AnypY27pPCJydHZPN2pFbkZGd1RKeU81OFY3ZHJldWc&output=csv", header = TRUE) # plot the data plot(mydata$u, mydata$v,, xlab="u", ylab="v") points(mydata$u,...
Ever wondered how to predict success or failure as a function of other variables? Here's a quick tutorial on binomial regression in R.
Which one of the sqldf, plyr, doBy and aggregate functions/packages would be faster for applying functions on groups of rows? I was wondering about this earlier in this post. It seems sqldf would be the fastest according to a post in manipulatr m...
Which one of the sqldf, plyr, doBy and aggregate functions/packages would be faster for applying functions on groups of rows? I was wondering about this earlier in this post. It seems sqldf would be the fastest according to a post in manipulatr m...
Heritage Health and Kaggle have teamed up to create the biggest data science competition thus far: the Heritage Health Prize, which challenges competitors to build a statistical model to predict the number of days a person is likely to spend in hospital over the next year, based on (anonymized) factors such as demographics, medical visits and treatments, and other...
In my last two posts I talked about Ordinary Least Squares, then extended this discussion to the multiple predictor case and briefly talked about some of the problems that may arise. These problems can include omitted variable bias, heteroskedasticity, non-normality, and multicollinearity. Most of these problems are relatively minor in practice and have easy fixes,...
In my last two posts I talked about Ordinary Least Squares, then extended this discussion to the multiple predictor case and briefly talked about some of the problems that may arise. These problems can include omitted variable bias, heteroskedasticity, non-normality, and multicollinearity. Most of these problems are relatively minor in practice and have easy fixes,...
Exercise 4 is all about using Newton's Method to implement logistic regression on a classification problem.
For all this to make sense i suggest having a look at Andrew Ng machine learning lectures on openclassroom.
We start with a dataset representing 40 students who were admitted to college and 40 students who were not admitted, and their corresponding...
Since I just didn’t get enough this morning, I spent some more time fooling around with estimating pi. Since I was basically counting the number of random x,y pairs inside a quarter circle and computing a sample average for more … Continue reading →
Zero Inflated Models and Generalized Linear Mixed Models with R.
Zuur, Saveliev, Ieno (2012).
