Bayesian methods in data analysis, an introduction
If you are in a sector where the outcome of your data analysis and machine learning work has significant monetory impact, then you should learn bayesian data analysis!
Bayesian methods in data analysis have been around for a long time. They are immensely helpful in solving complex decision analysis problems. Bayesian analysis is intuitively simple to understand and is computationally tractable
thanks to modern softwares like stan and pymc. On the other hand, they are rarely covered in introductory data analytics courses or even in engineering/college syllabus.
The purpose of this talk is to,
- Answer ‘why should you care about bayesian methods for data analysis?’
- Show their applicability and usefulness.
- Cover few interesting and fun examples (through code).
- Start with basics of bayesian methods, few historical anecdotes about the multiple interpretations of probability.
- Cover practical examples and problem statements which are best analysed with bayesian methods.
- Show some live coding examples using open source government datasets from fields like econometrics or agriculture or healthcare.
- Scratch the surface about algorithmic implementations: how the famous ‘markov chain monte carlo’ MCMC methods work.
- Quick review of libraries/tools (pymc).
- If you are excited with the idea, how can you study further?
A basic understanding of probability will help to understand the talk. The code examples will be in Python so some familiarity with Python is good too.
I work as head of data science at onlinesales.ai, an advertising technology startup based out of Pune. I have 7+ years of experience in data science and started in the field before it was a buzzword :-P. I have built multiple products, handled consulting assignments and delivered solutions using machine learning, R and Python. I hold a Master’s degree in Operations Research from Indian Institute of Technology, Mumbai.
Bayesian methods have been my area of interest for a long time. Over the years, I have formed few opinions about their usefulness and tried my best to understand the underlying theory, that I would like to share through this talk.