Review: Neural Network and its applications

Link to the paper

Why have I chosen this paper for a review?

The "Neural Network and its applications" paper has dragged my attention for several reasons. First of all, I have had prior experience with artificial intelligence and neural networks specifically, so I though that it might be interesting to obtain more knowledge in the field that interests me. Secondly, the structure of the paper appealed to me: I liked the way it was organized in multiple layers of topics and subtopics and liked the moderate amount of pictures (specially the kitty one in the bottom).

What did I not like?

I felt that the introduction was the weakest part of the paper. It is supposed to explain the most basic aspects of the subject and make me intrigued (or at least interested) in the topic. However, it left me clueless: I didn't receive any explanation about what neural networks and machine learning are, and why specifically the neural networks are interesting or important to study.

The biological part of the introduction appeared quite confusing. The reason might be the lack of my expertise in the biological field, but I faced too many unknown words and it felt like going too deep without getting to know the basics.

What did I like?

In my opinion, the topics are selected very thoroughly with the focus on the society-first applications of Neural Networks. Diagnosis, hazards, self-driving cars - these topics matter in the contemporary society. And DeepFakes as an insight into the future is a cherry on the cake.

My favorite part of this article is the DeepFake one, it is exceptionally well written and is interesting to read. I was always skeptical about Snapchat masks, deep faking and related topics, but I must admit that the paper has turned me more interested and beware of the dark side of the fun.

The part about the hazards has opened some new horizons for me as well, I was not aware of any nature related applications of machine learning except for weather forecast, but as it turns out its capabilities go beyond imagination. I believe, the hazard prediction is a very promising field.

Suggestions for improvement

Probably, to make this article more similar to actual wikipedia articles the authors could add more links to wiki pages for people unaware of such words as synapses, axons and similar (like me).

In overall, the article feels complete and informative, and I feel that I learned something new, which is the main purpose of wiki pages.