Sorry. Its been a little too long. I don't remember what thought about doing with it at the time. I do remember that I liked the idea, but could not get much response from the developer Steve
Slow? The software can analyze thousands of strategies a day. "3 days" refers to the time suggested a new user spend reviewing the tutorial. I'd like to know about software faster than Stratasearch.
Correct, It throttles your processors so a good cooling system is required but it's a workhorse that can prove useful.
i too am currently using the stratasearch free 30 day trial... but before i go off and tell you how complicated and confused i am using this software... i do think it has great potential... and THE MAIN REASON why i am confused... IS b/c I AM NEW TO TRADING... altho i have a firm grasp on strategies and indicators... i am not that ripe yet in giving a good synopsis on software so DO NOT take my word for it..... For me StrataSearch is confusing, but i do think it has great potential once i can sort out the cobwebs..... . Also, I followed the tutorial and out of the thousands of different strategies it produces, i dont know which one to use. Also you can input the dates you want analyzed for any particular sector and after it has run for a day or so it will produce a results list that you can view... this is where you should be able to decipher which trade setups and which strategies are the best... atleast that is what i think.... anyways, all in all, it is sooooo dam confusing b/c the size of calculations it runs that you'll spend more time deciding which calculated strategy to use then actually coming up with your own strategy. that is of course, if you can find the strategies that it analyzes! Stratasearch in a sense is a traders dream come true... it is by far the biggest work horse software out there... but its almost to big and complicated for its own good. they need to make it easier and less complicated to understand for us newbies.
Well, just about any software that uses a genetic algorithm for optimization. Using GA improves speed radically, and it does so as the number of parameter (indicators) rises.
QQQ,If i am not mistaken,I believe Stratasearch reccomends you run the system for 3 days,and possibly much longer...That was kind of a bummer for me... I will say that the program is far more sophisticated than I initially anticipated.Its a bit overwhelming at first,and its a bit unclear as to how one would actually use it.I think the people over at Stratasearch should consider a video of a step by step implementaion of the program.I for one would like to see how the developer intended the program to benefit traders. Its a bit difficult to get a feel for the program,and I actually stopped canned the demo after 4 days.I will revist it as i do think the program has great potential. Perhaps someone could tell me what benefit Stratsearch has over Nueral nets?? I am a curious novice.... And i there are any users out there who have actually enhanced their returns with StratSearch,I would certainly love to hear about it
Well Neural Networks discover solutions to problems more quickly than brute force methods. Genetic Algorithms are a better approach for optimization as they will scale better as you add more indicators/parameters. NN are more useful for modeling something like a stock chart to put together an idealized, tradable model of the stock. That is a simplistic analysis, but hopefully it gives you some ideas.
Most NN companies will tell you that NN solutions are within 10% of each other - meaning that they are close. Of bigger impact is the inputs you choose to put into the NN - that will have a much bigger effect than the NN itself. That said, some NNs do have advantages over other IMO - those that drive, consistently, to a single, repeatable answer (not true for all NNs) - tend to be better for trading than others. But I'll be the first to admit that I'm new to the field - so someone else may have a more detailed answer.
I have to say,after demoing StrataSearch for a week and really putting it to the test,I think it may be an excellent product.I am taking the liberty of quoting another user,but it is a very accurate description of its capabilities "StrataSearch offers varying degrees of automation for creating, testing and evaluating strategies. A brief summary of StrataSearch's automation layers follows: Strategies -- Manual build of trading rules, automated back test and manual review of test results AutoSearch -- Automated build of trading rules, automated back test and manual review of test results OneClick -- Automated build of trading rules, automated back test and automated review of test results Within each "layer", there is freedom to vary the degree of automation used. For example, Strategies can be setup to use only one parameter set ... or automatically vary one or more parameters within custom formulas. AutoSearch can be used to create strategies from randomly selected trading rules ... or build strategies using specified trading rules combined with randomly selected rules. OneClick can be modified to review results fed to it by custom built rules, randomly created rules ... or combinations thereof. The mix and match flexibility allows people to configure StrataSearch to fit their system development preferences. Naturally, it takes time to learn StrataSearch and get to the point where you can bend it into the shape you want." As i am finding out "People can strip out or add on layers of automation to help themselves find their "automation" comfort zone with StrataSearch.." I have no affiliation with the program,nor do i own it as yet,but I do think this is a program worth looking into...Just my 2 cents