# Home

Welcome to Likelihood Log

Likelihood Log is the hitchhiker’s guide to everything to do with probability theory, statistics, machine learning, artificial intelligence and Big Data.

This blog has two purposes.  The first is to monitor the pulse of data science and machine learning by selectively posting important, interesting or just plain odd news from the field.  The second, more ambitious purpose, is to provide an educational resource that helps experienced technologists get comfortable with the mathematics behind data science and, vice versa, helps pure math and stats specialists stay on top of technologies that drive modern data science.

I am a data science and artificial intelligence enthusiast. I am currently based in Sydney, Australia, working on several interesting data flow automation and predictive analytics projects.

When not crunching data, I spend time with family, try to stay fit, write code for fun, read about economics and history and host poker nights.

I am also an active networker and am always open to making new connections, receiving feedback and sharing knowledge. Please feel free to add me on LinkedIn, to follow on Twitter @likelihood_log or @LikelihoodLog or to browse my Facebook journal.

Disclaimer #1

In these posts, I aim to strike a balance between on one hand being detailed and technical enough to achieve structure and educational value, while on the other hand not getting too caught up in rigorous proofs and unreadable code, so as not to obscure the big picture “mind map” nature of this blog.

Thus, if you are an experienced software engineer you may find some of my code lacking in best practices, but hopefully the more math related posts will be interesting and challenging to you without drowning you in pedantic detail.  Similarly, if you come from pure mathematics or statistics background, then you may catch me cutting corners here and there in proofs or definitions for the sake of clarity, but then, you will hopefully be educated and challenged by the computer science and technology posts, all while finding the code readable.

Disclaimer #2

Likelihood log is not to be confused with log likelihood, which is the logarithm of the probability or probability density for the occurrence of a sample $\{x_1, ..., x_i, ...x_N\}$ given that the probability density $f(x;a)$ with parameter $a$ is known. In other words:

$\log L(a) = \log\Bigl(\prod_{\substack{1 \le i \le N}} f(x_i;a)\Bigr) = \sum_{\substack{1 \le i \le N}}\log f(x_i;a)$