He pasado la mayoría de la última semana construyendo recurrent neural networks a mano. Estoy tomando el curso de Udacity Deep Learning, y, llegando al contenido sobre RNN’s y LSTM’s, decidí construir algunos de ellos desde cero yo mismo.

When we look very closely at images generated by neural networks, we often see a strange checkerboard pattern of artifacts. It’s more obvious in some cases than others, but a large fraction of recent models exhibit this behavior.

Previously, I’ve written about the basics of scraping and how you can find API calls in order to fetch data that isn’t easily downloadable.

Here, we provide a brief introduction to reinforcement learning (RL) — a general technique for training programs to play games efficiently. Our aim is to explain its practical implementation: We cover some basic theory and then walk through a minimal python program that trains a neural network to play the game battleship.

Cross-posted at offconvex.org.

Although extremely useful for visualizing high-dimensional data, t-SNE plots can sometimes be mysterious or misleading. By exploring how it behaves in simple cases, we can learn to use it more effectively.

Below is the reference guide to Python that I wish had existed when I was learning the language.

El 2 de octubre del 2016, observé con terror como los colombianos votaron por el “No” en su plebiscito nacional para decretar el recién firmado acuerdo de paz. En la siguiente semana, medité sobre el resultado y las cosas que hubieran sucedido: la gran campaña de desinformación, las payasadas de Uribe y lo súper bueno que realmente parecía el acuerdo mismo. Hace dos días, me topé por casualidad con este post, que nos pide recordar que el margen escaso con el que ganó el “No” - 6,431,376 vs. 6,377,482 - no es particularmente convincente, ni, en realidad, tan decisivo frente al error humano.

On October 2nd, 2016, I watched in awe as Colombia’s national plebiscite for its just-signed peace accord narrowly failed. For the following week, I brooded over the result: the disinformation campaign, Uribe’s antics, and just how good the deal really seemed to be. Two days ago, I chanced upon this post, which reminds us that the razor-thin margin - 6,431,376 “No” vs. 6,377,482 “Yes” - is not particularly convincing, nor, as it happens, immune to human error.

A few weeks ago, I wrote an article about orbital mechanics and Lagrange Points. In my article I made the statement: