Author Topic: Q-Learning interesting Problem  (Read 597 times)

binharoun_7

  • *
  • Roomba
  • Posts: 1
Q-Learning interesting Problem
« on: November 14, 2013, 09:54:22 AM »
Hello everyone, i have just started to study Learning and see the possibilities of using Learning to solve my problem.

Problem: I am supposed to detect a certain combination of data, i have four matrices that acts as an input to my system, i have already categorised the inputs ( each input can either be Low (L) , or High (H) ). I need to detect certain types of input for example LLLH, LLHH, HHHH etc

NOTE:
1)LLLH means the first input in L, second input is L, third input is L and the fourth input is H!
2)I have labelled each type of input type as state, for example LLLL is state 1, LLLH is state 2, so on.

What i have studied in Learning is that most of the time you have one goal (only one state as a goal) which makes it easier for the agent to learn and create the Q-matrix from the R-matrix . Now in my problem i have many goal ( many states act as goal and need to be detected). I don't know how to design the states, how to create the Reward-matrix by having many goals and how the agent will learn. Can you please help me how can i use Learning in this kind of situation. Taking into account i have like 16 goals in 20+ states!

Freddy

  • *
  • Deep Thought
  • *********************
  • Posts: 5178
  • Mostly Harmless
Re: Q-Learning interesting Problem
« Reply #1 on: November 15, 2013, 04:42:50 PM »
Not something I know about. I can usually think of someone to ask but I draw a blank with this one.

Maybe you could try www.chatbots.org - the programmers tend to hang out there more.

I just wanted to say welcome too, so welcome :)

 

Welcome

Please login or register.



Login with username, password and session length
Picks Folley's
by Claude (Video)
Today at 03:26:32 PM
Ultra hal new programming language
by spydaz (UltraHal)
Today at 12:57:56 PM
DARPA creating software that updates itself (adapts)
by spydaz (General AI Discussion)
Today at 12:12:09 PM
What destroyed our world...
by Art (General Chat)
April 27, 2015, 11:16:46 PM
Action Scene Test
by Claude (Video)
April 27, 2015, 09:15:31 PM
Surfer
by Claude (Video)
April 27, 2015, 09:11:20 PM
Back from Prague
by ranch vermin (General Chat)
April 27, 2015, 04:33:59 PM
anyone want to have a chat or a stab at NLP with me?
by spydaz (AI Programming)
April 27, 2015, 06:47:01 AM
Video describes accelerating robot deployment in China
by Tyler (Robotics News)
Today at 10:51:09 AM
Han is a spookily realistic humanoid robot
by Art (Robotics News)
Today at 09:36:40 AM
Rethinking the Manufacturing Robot
by Tyler (AI News )
Today at 04:50:30 AM
Reddit AMA with Robohub team this Tuesday April 28 starting @ 2pm EST
by Tyler (Robotics News)
Today at 04:50:29 AM
One way to reduce email stress: Re-invent the mailing list
by Tyler (AI News )
April 27, 2015, 10:53:48 PM
The Drone Center’s Weekly Roundup: 4/27/15
by Tyler (Robotics News)
April 27, 2015, 10:53:47 PM
Japan looks to distributed, cooperative control to help manage energy market deregulation
by Tyler (Robotics News)
April 27, 2015, 04:50:35 PM
Parents sound off on mobile device use by children
by Tyler (Robotics News)
April 27, 2015, 04:51:36 AM

Users Online

17 Guests, 1 User
Users active in past 15 minutes:
Claude
[Trusty Member]

Most Online Today: 41. Most Online Ever: 208 (August 27, 2008, 08:26:54 AM)

Articles