A/B Tests and Experiment Size
Let’s say you’re running an A/B test. Maybe you want to test how many conversions you will get if you change the design of the Signup page or the wording. Yo...
Posts by Category
Blog
NLP Pipelines with NLTK
Often with Natural Language Processing (NLP) applications a pipeline is useful to take the raw text and process it and extract relevant features before input...
Useful Pandas Snippets
Imputing missing values
Deploying imgproxy with AWS Fargate
imgproxy is a fast and secure standalone server for resizing and converting remote images. The main principles of imgproxy are simplicity, speed, and securit...
The Beta Function
In Bayesian statistics, often the Beta function is used as the prior distribution for some unknown parameter in a Bernoulli experiment. In this post we compu...
Mathematics of the t-test
There are times when our sample size is too small for asymptotic results to be valid. We cannot simply use Slutsky’s theorem to replace the true variance wit...
Method of Moments
Definition
Fisher Information
Log-Likelihood
Vectorized Backprop: Coding it up
In a previous post, I walked through the maths of back-propagation (“backprop”). Here I will go through the implementation in Python (heavily based on Andrew...
Vectorized Backprop: The Maths
Design matrix
Forecasting the price of bitcoin using an LSTM
I’m sure that every developer who has learned a little bit of ML dreams of applying this to the stock market and getting rich. At first glance RNNs and LSTMs...
blog
Setting up Jupyterhub on AWS
This guide will be about setting up the fiddly bits when deploying a Jupyter Hub to an AWS instance. It won’t go into explicit detail about absolutely every ...
Setting up a public Jupyter Notebook on AWS
Jupyter Notebooks are a great and widely used tool in data science. Quite often then run are run on localhosts or have to be accessed via SSH tunnelling.
COVID-19 Analysis: what can the data tell us?
This year has been like no other in living memory. Coronavirus has shaken our collective sense of what normality means and will undoubtedly continue to affec...
Coin flip power analysis
Deciding if a coin is fair
Multidimensional Central Limit Theorem
Instead of just having a single random variable $X$, we may have an experiment for which we are recording several random variables, which we can consider as ...
The Multivariate Delta Method
Building on the univariate delta method
The Delta Method
The Delta Method
Django database optimisation
The Django ORM makes interacting with the database a breeze, but without due care can also lead to poor performance.
Sample variance
Where does the mysterious $(n-1)$ factor come from when computing the unbiased sample variance?
Kelly Criterion
If we bet a fraction $f$ of our capital on each round, and a win resulted in a gain of $f\times b$, and a loss resulted in a loss of $f \times a$, then after...