OpenAI has been on the news for some time. The media attention of OpenAI is deeply related with the interest in Reinforcement Learning (RL) applications.

In the graph below we see the google searches related with RL and OpenAI. That is not a coincidence since OpenAI offers a tool used by many in the field, which is the OpenAI Gym.


Monte Carlo Tree search is a fancy name for one Artificial Intelligence algorithm used specially in games. Alpha Go reportedly used this algorithm with a combination of Neural Network. MCTS has been used in many other applications before that.

Here I explain what algorithm is, and how it works.

What is Monte Carlo Tree Search

MCTS, like the name says, is a way of searching a tree. Within this tree, its nodes represent the states and the arc between nodes represents the choice that would take from one state to the other.

The image below shows the game of Tic-Tac-toe and how we can represent its…


PyTorch is an AI framework developed by Facebook. Its code is available on GitHub and at the current time has more than 22k stars. It has been gaining a lot of momentum since 2017 and is in a steady adoption growth. To understand this trend more and put it in perspective, we can see the picture below containing Google trends data comparing the searches for Tensorflow and PyTorch.


I tried introducing how to deploy an AI application. Please leave a comment about it or about something else that you would want me to write next. Follow me and show some love clapping if you like it!
LinkedIn
Twitter
Github

You created your Deep Learning model. You had excellent results in your tests. But what do you do with this model? How would you allow other people to use it?

The most common answer to this question is that you want to create a web server to let others use your model over the internet. This can easily be…


I tried introducing how learning happens on a Neural Network. Please leave a comment about it or about something else that you would want me to write next. Follow me and show some love clapping if you like it!
LinkedIn
Twitter

You started working with Deep Learning. You have a problem to solve. You go and choose a framework like Keras, TensorFlow, Pytorch. Then you build your network, give it examples and let it learns. Magically, it does really learn. But what does it mean to train a network? What is it performing under the hood?

The short answer for…


I tried giving the intuition on how Style Transfer works. Please leave a comment about it or about something else that you would want me to write next. Follow me and show some love clapping if you like it

Style Transfer is an application of neural networks that is fairly recent. It gained its popularity by showing outstanding transformation of photos based on artist paintings. The original work is described in Gatys et al (2015).

The image displayed below shows several examples of this paper where a picture is transformed to something that is painting like.

Here I try to…


I tried giving the intuition and basic components of how CNNs work. Please leave a comment about it or about something else that you would want me to write next. Follow me and show some love clapping if you like it

Convolutional Neural Networks (CNNs) are the Neural Networks tools for problems related with images. It can also be used in other domains, but it is responsible for the big leap in advancements with everything related to image processing. It is part of several camera related programs we use every day.
For instance, It is used for detecting a human…


Recurrent Neural Networks (RNNs) are the Neural Network tools for problems that deal with sequential data.
It became increasingly more popular due to their great results in Natural Language Processing (NLP).
Within NLP they are used for the most varied tasks like translation, text classification, automatic text generation.

Here, I briefly explain the basic structure of RNNs

Basic Building Block

RNNs are normally shown like the picture below

Representation of RNN both in folded and unfolded forms

In the picture, there are some distinct components, from which the most important are:

  • x: The input. It can be a word in a sentence or some other type of sequential data
  • O: The output…


Serveless infrastructures are the solution for implementing true auto-scaling and tying your costs with how much your application is used.

The traditional path of putting software into production is to test on a development computer and then try to find a server to host your code. Sometimes it can be a server maintained by the organization itself. Where you buy all the hardware and install the software there.
What has been very popular is to hire cloud resources from companies like Amazon, Google, IBM and so on to run your code.
Normally these hired services charge by the hour that you…


Upverter is an entirely online Printed Circuit Board software. I have used it since around 2016 to design something like 4 boards. I am not associated with Upverter in any form, but I do believe it is a very useful software worth mentioning.
In this article I am going to briefly discuss about its characteristics and my personal opinions.

Pricing

Upverter is currently Free.
This aspect is specially important if you want to develop commercial products. It is common that companies offer a limited free version of a software where you could only use it to open source/open hardware products.
In…

Pedro Borges

MS in CS — UCLA. BS in Mechatronics Engineering — UnB. pedro-torres.com linkedin.com/in/pedrohbtp

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