neural networks for predecting the stock market ..... good one

what do I set for learning rate 0.1
momentum 0.01
the values are from 0-100.

Here we are, I look for somebody to learn me to use the software of NeuNet Pro in his(her,its) global nature and also I interresse me in horse-races.
Thank you in advance for your answer and for your help(assistant).
I live in Paris.

Hi,

I am very new to the area of neural networks and was hoping somebody would be able to give me possible suggestions on how to design the neural network for my particular problem.

Any sort of help would be greatly appreciated.

1. I have a large continuous action and state space.
2. My goal is to find a good action to take while in a particular state.
3. However, I do not have any training data. I do however, get to know the cost of taking an action in a particular state after of course, taking the action.
4. Hence, I was thinking of using this cost somehow to train the net to choose actions when in a state. I could perhaps, train the net by adjusting the weights appropriately, when the cost goes up or down from the previous time period.

Most of the research seems to be for supervised learning with training data. However, I have not been able to find papers that want the neural network to learn a mapping from state to action in a continuous and large action space.

Any suggestions?
Thanks
Annapurna

after my previous post, a thought occurred to me - does the dll require that a '.net' file be closed before another is opened? This could (I think) explain it - the program will be spending most of its time opening and closing the '.net' files, even though they are on a memory disk. If this is the case, is there any way round it?