You can find many webpages which are focus on the research work of machine learning, e.g., NLP and CNN.

As you know, the success of AlphaGo over the human beings has proven that the computer can solve a strategy problem like the human.

We have built many robots in the past years, but these robots are not intelligent; they can execute many actions under the design of humans.

We can say that these robots are lack of brains which are intelligent enought for them to finish the tasks under the order of humans. In other words, human just need tell these robots what should be done while never care about how to achieve these goals.

We need the robot act as the servants, rather than just the machines which are powered by electricity.

I think next time, I need publish researh papers more, because they are my research works.

Therefore, we believe that the combination of deep learning and reinforcement learning can achieve such a dream.

In many papers such methods are called Deep Reinforcement Learning.

We think such a work is related to the research field of Artificial Intelligence Automation.

I hope that you can find something useful here, if you are interested in the topic of Artificial Intelligence Automation.

I hope that I can publish a book in the future. However, you can translate it into Chinese.

There is a paper are prepared for you!

A Design Method of Fuzzy Logic Controller by Using Q Learning Algorithm

http://xueshu.baidu.com/usercenter/paper/show?paperid=3d340b8697d7ffadaac7c806792b1841

Let the voice be listened by all over the world.

and there is a group named AAIA

If the target positions are setted randomlly, it if hard for the Reinforcement Learning (RL) algorithm to achieve the task.

Therefore, I cannot say the RL is intelligent enough.

There are some details about the intelligence of human are not discovered presently.

The success of RL agent in the control of robot which has continous action space.

The parallelization of RL has already been apllied in many research works.

Maybe I should use A3C algorithm rather than the NAF or the DDPG.

Artificial Intelligence Automation. Every day.