In part 1 we got to know the openAI Gym environment, and in part 2 we explored deep q-networks. We implemented a simple network that, if everything went well, was able to solve the Cartpole environment. Atari games are more fun than the CartPole environment, but are also harder to solve. This session is dedicated to playing Atari with deep…
Introduction to OpenAI gym part 2: building a deep q-network
In part 1 we used a random search algorithm to “solve” the cartpole environment. This time we are going to take things to the next level and implement a deep q-network.The OpenAI gym environment is one of the most fun ways to learn more about machine learning. Especially reinforcement learning and neural networks can be applied perfectly to the benchmark…
Getting started with OpenAI gym
The OpenAI gym environment is one of the most fun ways to learn more about machine learning. Especially reinforcement learning and neural networks can be applied perfectly to the benchmark and Atari games collection that is included. Every environment has multiple featured solutions, and often you can find a writeup on how to achieve the same score. By looking at…
The Zalando python client
The Zalando python client Shopping is probably in the top-5 of things I hate. Walking from store to store, having to wade through endless clothes you are NOT interested in – it’s a really inefficient experience. Luckily I found Zalando a few years ago: you order your clothes online with free shipping and free returns. The only thing that could…
Simple Introduction to Tensorboard Embedding Visualisation
Visualising embeddings is a powerful technique! It helps you understand what your algorithm learned, and if this is what you expected it to learn. Embedding visualisation is a standard feature in Tensorboard. Unfortunately many people on the internet seem to have some problems with getting a simple visualisation running. This is my attempt at creating the most simple code to…
Laser engraving a wine box
This weekend my dad turns 63. This year I created a very personal gift for him: a laser engraved wine box. Recently my local library opened a fab lab (fabrication laboratory). For only a few euros you are allowed to use the machines they have. With three 3D printers, a laser cutter, and a dye-sublimation printer, children can learn a lot…
What I learned developing the Neural Voting Advice Application
Two weeks ago I put the neural voting advice application online (http://www.pinchofintelligence.com/neural-voting-advice-application/). In this post, I look back at two aspects: the end-user aspect and the technical aspect. The first one will be interesting for people working with neural networks. The technical aspect might be useful for somebody else interested in deploying their Tensorflow or Keras application. Introduction For years…
Stemwijzer met kunstmatige intelligentie
Volgende week zijn er verkiezingen in Nederland. Met 28 partijen is dit een lastige keuze. Vaak orienteren mensen zich door een stemwijzer in te vullen. Door de 30 belangrijkste vragen van het moment in te vullen zoeken ze een partij die het beste bij ze past. Stemwijzer met kunstmatige intelligentie Al jaren denk ik dat dit beter moet kunnen. In…
Neural voting advice application
Next week the Netherlands will hold elections to choose who is going to represent the people in the house of representatives. There are 28 political parties participating, although there are 10 ‘big’ political parties most people choose from. As there are so many political parties people orient themselves on who to vote for using voting advice applications. Most of the…
Watching the TensorFlow Dev Summit 2017 – Livestream
Last night Tensorflow held the Dev Summit 2017. Although it was in California, they streamed all talks live through Youtube (https://www.youtube.com/watch?v=LqLyrl-agOw). As I use Tensorflow a lot I loved watching some of the talks (together with 2500 other people watching live). This blog post is a summary of the first talks I watched before going to bed. First of all,…