Learning about Machine Learning & Artificial Intelligence
A beginner to practitioner guide
“What I learned on my own I still remember”
N.N. Taleb
Why should you learn it?
Data Science has been disrupting the medical, businesses, politics and socio-economic sphere in the last decade by automating and augmenting, mainly from the most mundane to even expert-level systems. Beyond its buzz, there are many everyday examples of Artificial Intelligence sprouting all around us without even us noticing any of it.
It seems appealing to many software engineers and web developers to start acquiring those highly desirable skills to be ready to apply Machine Learning. But they are usually daunted with the sheer amount of pre-requisites to start learning.
What all do you need to learn?
These state-of-the-art techniques help drive business decisions has its roots deeply embedded in methods ranging from Bayesian Inference, Linear Algebra, Statistical Modeling, Natural Language Processing, Convex Optimization, Data Wrangling, Database Management, Information Retrieval, Big Data and such. So it would be great to review those concepts before diving into the realm of Data Science.
Data Science is a highly interdisciplinary area, where it requires knowledge of four primary domains:
- Hacking Skills
- Data Visualisation
- Higher Mathematics
- Business Acumen
Thus, knowing R or learning a bunch of python libraries would only improve your hacking skills and would require a lot more efforts in the other domains as well.
Be advised! There exists no one-size-fits-all, cookie-cutter model which will solve your business problem. Hence a broad understanding of various techniques & models would require a lot more efforts.
Where can you start learning it from?
Most of my Data Science and KDD resources include Shapiro’s KDnuggets and India based Analytics Vidhya. They both have plenty of resources for getting you started to find a job in Data Science.
Including links and information to great primer course for all ML newbies, Andrew Ng’s course for Machine Learning and the latest course for Deep Learning as well.
After having enough confidence to apply your skills you can move over to Kaggle and pick up a challenge to solve.
Every developer’s good old friend Stack Overflow has these helpful sister sites which are domain-specific.
Once engaged in the struggle and deploying production level predictive models, you will understand the misconceptions surrounding the buzz and learn how to communicate efficiently while producing cohesive results for driving the business decision forward using data.
Appendix
- History behind term Data Mining
- Andrew Ng’s answer to How can beginners in machine learning, who have finished their MOOCs in machine learning and deep learning, take it to the next level and get to the point of being able to read research papers & productively contribute in an industry?
- Artificial Intelligence is a total misnomer — Manuel Ebert — Medium
Happy Learning!