2. About

About

Motivation

Number of ML related information sources is unleashed. In Michał Chromiak on Machine Learning (MCML) blog I aggregate and synthesize the most important and valuable knowledge on the topic. From time, to time, I also write some article on topic that I subjectively consider interesting. What drives me into ML is its power to "cast mathematic spells" on data. I got inspired to pursue towards the ML by people like Andrew Ng or Fei-Fei Li who not only are experts in ML, but also makes all the best efforts to democratize this area.

It is fascinating how AI becomes an art of modern days. Its crux is automation, mainly automation of processes with ML. As of today I believe that DL is currently the brightest past for future of ML. AI expert should be a polymath to embrace the concept of mind. The AI brings together the philosophy, psychology, neuroscience and neurobiology.Being a modern AI expert is like being a science artist who knows how to optimally apply AI to a problem. I am fascinated by the AI and its possibilities. This blog aims to contain some amazing ideas that I have approached considered special :)

Code

Some nice pieces of code I keep in GitHub ML-Kraft and the learning materials on ML tools and frameworks in GitHub ML-DOJO.

Bio

I am an assistant professor at the Maria Curie-Skłodowska University. I hold a Ph.D. in Computer Science from Institute of Fundamental Technological Research, Polish Academy of Sciences (PAS). My thesis involved creating unified integration platform for heterogeneous resources. Along the PhD research I have also participated in research on Stack Based Approach to databases, at Polish-Japanese Academy of Information Technology developing prototype of ODRA object database. Additionally in 2016 after Ph.D. due to being fascinated by the CRISPR revolution I have done some research on modeling the CRIPS workflow with map-reduce framework.

Over the course of my PhD I squeezed in couple of industry roles like development of aerial image georeferencing platform for IUNG or devising a transaction-alike system based on Elasticsearch for CompuGroup Medical.

Desiderata

I believe that willingness to be a fool is the precursor to transformation. Thus, learning require you to be a fool, first. Here is where I put my notes on understanding the nature of machine learning right from the perspective of a fool - that at the beginning I was/am (depending on perspective of what I'm learning).

After all, first you need to be a fool... to start the endeavor and to be in-formed of the new.

I will try to explain things that I learn however, everything can be explained in two ways:

Look how clever I am

or

Look how easy it is

I always do my best not to get tempted by the first option ;) If you find the explanations to complex or find an error, please let me know what you think.

Please feel free to contact me if you have a suggestion on present content, or would like me to write on some specific topic. I am always also open to work on some nifty new ideas.

P.S.

Before you start to read an absolute must for start is to handle couple basic things:

  1. Intermediate Python experience.
  2. Multivariable Calculus and
  3. Linear Algebra

To setup your development environment:

  1. Intro to Data Analysis introduces Numpy, Pandas, or Matplotlib which are main tools for working with and visualizing data in Python