Meor,

> I wanted to share about another project I've just started called
> kDimensions. It uses visuals to explain machine learning concepts.

Thanks for sharing your material. One of your diagrams distinguishes
between "rule-based systems" and "machine learning", but there has been
plenty of work on machine induction of rules, some leading to deployments.

The same diagram also suggests that machine learning relies on large
data sets, but there are many methods that extract expertise from small
training sets. Again, some of these have produced fielded systems.

If you would like to read about some early results along these lines, you
can find an early review article here:

Langley, P., & Simon, H. A. (1995). Applications of machine learning
    and rule induction. Communications of the ACM, 38, November, 55-64.
    http://www.isle.org/~langley/papers/mlapps.cacm95.pdf

The popular rhetoric is that the successes of machine learning have all
been recent, but the field has been producing useful results for decades.

Best wishes,
Pat Langley