Scribd
Upload
387 views1 page

Think Like A Machine

DeepMind is a leading artificial intelligence research company that was founded in 2010 and acquired by Google in 2014. One of DeepMind's focuses is reinforcement learning, and they developed an innovative method called DQN using deep neural networks and the Q-function. DQN was able to outperform other AI techniques by becoming a game tester itself and beating human game players. This chapter discusses using a similar approach to create an intelligent adaptive network for optimizing transport tasks in warehouses using reinforcement learning.

Uploaded by

Rady
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
387 views1 page

Think Like A Machine

DeepMind is a leading artificial intelligence research company that was founded in 2010 and acquired by Google in 2014. One of DeepMind's focuses is reinforcement learning, and they developed an innovative method called DQN using deep neural networks and the Q-function. DQN was able to outperform other AI techniques by becoming a game tester itself and beating human game players. This chapter discusses using a similar approach to create an intelligent adaptive network for optimizing transport tasks in warehouses using reinforcement learning.

Uploaded by

Rady
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 1

Think like a Machine

DeepMind is no doubt one of the world leaders in applied artificial intelligence.


Scientific, mathematical, and applications research drives its strategy.
DeepMind was founded in 2010, was acquired by Google in 2014, and is now
part of Alphabet, a collection of companies that includes Google. One of the
focuses of DeepMind is on reinforcement learning. They came up with an
innovate version of reinforcement learning called DQN and referring to deep
neural networks using the Q function (Bellman's equation). A seminal article
published in February 2015 in Nature (see the link at the end of the chapter)
shows how DQN outperformed other artificial intelligence research by
becoming a human game tester itself. DQN then went on to beat human game
testers. In this chapter, the agent will be an automated guided vehicle (AGV). An
AGV takes over the transport tasks in a warehouse. This case study opens
promising perspectives for jobs and businesses using DQN. Thousands upon
thousands of warehouses require complex reinforcement learning and
customized transport optimization. This chapter focuses on creating the reward
matrix, which was the entry point of the Python example in the first chapter. To
do so, it describes how to add a primitive McCulloch-Pitts neuron in TensorFlow
to create an intelligent adaptive network and add an N (network) to a Q model.
It's a small N that will become a feedforward neural network in Chapter 4,
Become an Unconventional Innovator, and more in Chapter 12, Automated
Planning and Scheduling. The goal is not to copy DQN but to use the conceptual
power of the model to build a variety of solutions.Think like a Machine Chapter
2 [ 29 ] The challenge in this chapter will be to think literally like a machine. The
effort is not to imitate human thinking but to beat humans with machines. This
chapter will take you very far from human reasoning into the depth of machine
thinking.

You might also like