IdeaCoding Lab
Published in

IdeaCoding Lab

“We believe deep learning and JavaScript belong together”, me too!😍🤗

“You’re mine
And we belong together
Yes, we belong together
For eternity” We Belong Together

Photo by Andres Molina on Unsplash

“We believe deep learning and JavaScript belong together”, said Nielsen et al

Me too! 😍🤗

I have to be totally honest, the scene is still dominated by Python: saw that on my postdoc, saw that on a live I did months ago with a researcher working with cutting-edge research on bioinformatics.

Asked him regarding this possibility to use JavaScript, and his answer was regarding the packages already being in Python; saw that when I was working with Java, and most were working with C++, can be quite a challenge to keep up.

The good news is they build TensorFlow.js to be competible with TensorFlow, it seems you actually can transform the packages, and use predefined neural networks.

it is possible to transform models in both directions

Here goes some strong points to consider regarding JavaScript. I would go even deeper: Angular can be even better, since it has services and their logic is separated from the User Interface. Recently, I have been adapting JavaScript codes to Angular, a quite an interesting experience, I have to say!

JavaScript is present easily to anyone, no maintenance

“Thus, should you be in doubt, it is true that python is strong for desktop applications, but, consider the following reflection: are we still desktop applications driven? I think not, I think web applications are the future, and so it is JavaScript and any related technologies, such as TensorFlow.js.” My selected assays from Medium on Computer programming

One of the biggest weakness of deskop application is maintenance. The ones that are old enough to remember, we needed to constantly update the software, creating costs and stress not just for us, but also for the providers. Today, it happens before you see it, automatically.

Integrating information can be easier

“One think people tend to forget when dismissing computers as their companion, e.g., medical doctors, is their ability to integrate information: fast, efficient, and cheap. With proper care, also reliable.” Computer thinking, my new book coming out, keep your eyes sharp here 👀👀👀

MEAN stack is quite powerful, even more power when using NestJS. MongoDB is growing like crazy, used on my PhD, and was quite impressed by its power for handling big data.

Final remarks

I will leave fuel for other articles, but I am quite excited about this marriage. Of course, just time can tell, as said Enya.

Q&A: based on online questions and comments

A weak typed language with deep learning, do not see it going well

First of all, I cannot see the correlation between deep learning and typed language, they are too different concepts, as I see it. One is artificial neural networks theory, machine learning theories. They are independent of what language you use, and suppose to; think them as pseudocodes, deep learning is an algorithm. Imagine a theory of machine learning that depends on the language, sounds useless to me, and this is not the case of deep learning.

Second point, according to the authors of the book mentioned, you can exchange codes between python and TensorFlow.js. They claim to have solved all the issues. TensorFlow.js is behind several Google Technologies: Firebase, AutorDraw and more.

Third, you do can use TypeScript, without too much of an effort.

I guess that close the case!

Jorge Guerra Pires, PhD, site

Independent Researcher / Udemy Instrutor/ Superprof Star professor [PT] [EN] / Independent Writer at Amazon / CV Lattes

Member of the Center of Excellence for Research DEWS (University of L’Aquila, DISIM, Italy)

About me, short bio

I have been working with computer programming and mathematical modeling applied to biological systems since my bachelor of Engineering.

Currently, I am an Independent Researcher and Member of the Center of Excellence for Research DEWS (University of L’Aquila, DISIM, Italy). Love writing and reading!

Short-bio: B.Eng by Universidade Federal de Ouro Preto (Brazil) in Production Engineering; double master degree by University of L’Aquila and Gdansk University of Technology: PhD on a collaboration between the Brazilian program Science without Borders in biomathematics at the University of L’Aquila/IASI-CNR/BioMathLab. Master of science and PhD degrees recognized in Brazil by University of São Paulo (USP) as bioinformatics. Postdoc by Federal University of Bahia and Fiocruz.

--

--

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store