Sadly this is another post aimed at R users. Recently I’ve been looking into how to make my programming cleaner, one of the main ways of doing this is a method called Object Orientated Programming. It’s used by most of the functions in R for example look at the output from a linear model.
#Fit linear model lm_cars <- lm(mpg~.,data=mtcars) is.object(lm_cars) ## [1] TRUE isS4(lm_cars) ## [1] FALSE class(lm_cars) ## [1] "lm" So the first check in the example above checks if lm. »
MACHINE LEARNING!!! So, if you’ve not heard of machine learning yet, you probably haven’t been watching any TV the last decade. Problem is machine learning is absolutely massive field.
This is the intro to a series of blog posts I plan on doing on various areas in machine learning, more specifically statistically backed methods therefore I call it statistical learning. In this blog post I aim to break down the two main areas that are generally focused on. »
So, I’ve upgraded from the old and outdated ad-riddled wordpress, up to a nice hugo website generated by blogdown, an R package developed by Yihui Xie. You can have a look at my old wordpress here. In my opinion the new blog is a huge improvement and much more efficient too. Here I’ll give some basics on how to get started on making a similar one.
So why did I update? »
Last week I attended a talk given by Professor Idris Eckley named “Making Sense of Sensor Data”. One of the main areas I see sensors being used every day is my phone telling me how much I’ve walked. There is quite a few problems in sensor data. I’ll look at two of them here namely the changepoints in the data and the frequency at which we observe data
Changepoints The first is quite simple, suppose we observe a long uni-variate time series of some movement measure in a sensor. »
If you can’t guess from the title this post is about Extreme Value Theory (EVT). Last week I attended a talk from Professor Jonathan Tawn on some EVT applications and Jenny Wadsworth on some theory in multivariate EVT. I found it particularly interesting as unlike most methods of analysing data we don't really care about the data which isn't surprising, instead we focus on modelling the tail ends and unlikely observations. »
Welcome to my Blog! I am an MRes student at STOR-i, Lancaster University and as part of my development as a student I will be maintaining a blog. STOR-i is a EPRSC funded Centre for Doctoral Training (CDT), it mainly focuses on applications of STatistics and Operations Research in Industry. To read more about me you can click on the STOR-i Website button in the menu.
Anyway about the actual blog. »