Learning Notes · Experiment: Learning reflection · Deep Learning Specialization · Short read

Learning Deep Learning by Explaining It

By Obada Alsehli

A personal reflection on using equations, explanations, comparisons, and retrieval questions to build a reusable learning practice around a deep-learning course.

The notes were part of the work

When I shared completing the Deep Learning Specialization, I wanted to talk about more than the certificate. Alongside the course work, I wrote structured reference notes on the concepts I was studying.

Those notes included equations, gradient derivations, architecture comparisons, common mistakes, and questions for retrieval practice. The aim was to create something I could return to, rather than a summary that only felt familiar immediately after a lecture.

Explaining a mechanism exposes a gap

Writing an explanation asks more of me than recognizing a term. To explain a gate equation, I need to account for the terms and the behavior they produce. To compare architectures, I need to state the difference that matters, not only repeat their names.

If I can follow an explanation but cannot reconstruct it, that is useful information about what I still need to study. A reference page becomes more valuable when it records that uncertainty instead of hiding it behind a fluent paragraph.

This was the learning practice I described in the original post. It is a personal account, not a measured comparison of study techniques or a claim that note length proves mastery.

A certificate and a learning system have different jobs

A credential records completion. A set of explanations, questions, and worked reasoning can support the next problem after the course ends.

For me, the important habit is to move between studying a concept, trying to explain it clearly, and identifying what I cannot yet justify. The point is not to document everything; it is to make the understanding inspectable enough to improve.

That is the part of the specialization I wanted to keep building on: not just having seen the material, but having a way to return to it and think with it.

Credentials archive

Cite this work

Obada Alsehli··n.d.

@article{alsehliwritinglearningdeeplearningbyexplaining,
  author       = {Alsehli, Obada},
  title        = {{Learning Deep Learning by Explaining It}},
  howpublished = {\url{https://obadaalsehli.com/writing/learning-deep-learning-by-explaining}}
}

Academic attribution & reproducible reference format

BIBTEX