Connect with Facebook
Log in |   Register | Guidelines | Search | |


Long Term Trade Recommendations Recommendation are given by members to members and are not endorsed nor warranted in any way by MoneyTec. All members participating are responsible for their own statements.

Reply
  #151 (permalink)  
Old 04-24-05, 06:02 PM
Batcheler's Avatar Batcheler Batcheler is offline
level 3
 
Join Date: Jun 2004
Age: 60
Posts: 327
Thanks: 0
Thanked 0 Times in 0 Posts
Rep Power: 9
Batcheler is on a distinguished road
Re: hi good folks

Quote:
Originally Posted by bigcat
Hi Jerry,

Now I think I see, thank you for your last explanation. Those SUPER signals were uncannily accurate ... and now I see we would've waited and waited and not seen one in real time.

Your idea in the 2nd paragraph sounds pretty good! When the market is choppy, it doesn't seem like much would be lost entering a day or two later. When the market is trending, it is true we would likely have a worse entry, BUT on the other end sometimes we will get a better exit (see the second to last buy signal on the EUR chart -- had it been delayed, the SHORT we were in would've run further).

I hope you will keep us updated on the current status!

Thanks,
Bigcat
Thank you, Steve, for the "blast of refreshing fruit flavor for your mouth!" (old commercial), LOL, i.e., the school teacher lecture. we know this stuff, but thanks anyway Steve, always good to keep our noggins in those goods! As if, too, anyone is perfect enough to not get caught on "if its too good...." EVER! If so, raise your hand, wanna see you walk on water too! LOL i.e., let's keep this thread as helpful members in this founder's eNNya group, otherwise, I'd say finger-pointing and nasties added into the weathering I'm going thru......why keep thread here if such?

Good point bigcat, just ran both EUR & JPY models (recovered those eNNya-II models) and they are both just now in a SL. Now to see what if any Sigs come next, so we don't know what will happen, but you made good point, where we lose at such delay the some other times we'll gain back. And besides, then too that I over-looked in my blow-up haste is that there are trades in between those delayed trend points that eNNya-II DOES keep up with in the interim!
Sponsored Links

  #152 (permalink)  
Old 04-24-05, 09:21 PM
Batcheler's Avatar Batcheler Batcheler is offline
level 3
 
Join Date: Jun 2004
Age: 60
Posts: 327
Thanks: 0
Thanked 0 Times in 0 Posts
Rep Power: 9
Batcheler is on a distinguished road
Good News!

Good News!

Did first Trend peak walk around of peak-spike at 2/10/05 of USDJPY chart, did walk arounds of 2/10, 2/16 & 2/23. Short cuts-out thus Long trade early at 2/3, finally it appears ON THE CHART EARLIER somewhere between last two dates, so it's like having, in some cases, an earlier trailing_stop, or simply a cut-out of trades based on eNNya's other TAs in the hopper. From first of the year I see equities around 550 pips, is that X.7 for USDJPY for $$ gang? Anyway, it's looking like not the end of the world, thank God, for now. In the range of week-'n-half later (between last two above dates) different Sig suddenly appears off to left of the hump (as surmised). If still in opposite trend on right side seems we could go ahead with that direction-trade too, until we find such plans don't work, but if eNNya says after post AutoTrend indication re-train adjustments that we're still in that inclination, seems ok from here. Nice part about the Trends coming into the equation and doing re-train is the blue-line is kept in view for the times between and during trends, so all in all, things looking not too bad.

It's a painstakingly slow tedium process to look this way around points of interest with such an anomalie, new datasets for each end-date, new data-manager entry, new chart, new model, on' n on, for each date of interest! But I got one looked at for now that yields a "whew!", probably can get in couple more over my Mon-Wedn weekend. OK? Don't think it was a killer as it seemed before. It was such a slow rising to the surface "hmmm", followed by a bang at end of digging for what's happening.

I'm happy to you folks that you are hanging in there with me. This IS certainly an engineers task, it's NO WONDER HUGE percentage don't make it doing this stuff with fancy softwares/NNs, it takes an engineers regimen, to wit even the "hmm"-to-digging-to-"whats this?" TOTALLY re-surfaces my engineering days in electronics! And I'm GLAD for such because menial jobs approaching Social Security in 6 years is SO boring! So this is actually real life lifter for me, real confidence lifter. (not to mention our eventual going forth with THE LOOT! $$$ Ahhhh (scream) LOL
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!Spurl this Post!Reddit! Wong this Post!Twit this!
Reply With Quote
  #153 (permalink)  
Old 04-24-05, 09:57 PM
Batcheler's Avatar Batcheler Batcheler is offline
level 3
 
Join Date: Jun 2004
Age: 60
Posts: 327
Thanks: 0
Thanked 0 Times in 0 Posts
Rep Power: 9
Batcheler is on a distinguished road
more in this venue!

more in this venue of good news!

thoughts that during eNNyas-II's early cut-out of trends we could have eNNya-I show us what she thinks of the potential or actual peaks. (re-waken her during these moments) Would mean back to 5:30 & 6pm Sigs to do both at those times, but could be worth it!
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!Spurl this Post!Reddit! Wong this Post!Twit this!
Reply With Quote
  #154 (permalink)  
Old 04-24-05, 10:34 PM
bigcat's Avatar bigcat bigcat is offline
level 1
 
Join Date: Apr 2005
Posts: 21
Thanks: 0
Thanked 0 Times in 0 Posts
Rep Power: 0
bigcat is on a distinguished road
Re: more in this venue!

Aren't second opinions usually a good idea, when risk is involved? It is interesting that eNNya-I might serve to offer a second opinion, when needed, for eNNya-II.

It seems like the last round of "troubles" is working toward making the approach stronger. Hang in there Jerry!
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!Spurl this Post!Reddit! Wong this Post!Twit this!
Reply With Quote
  #155 (permalink)  
Old 04-25-05, 05:21 AM
stevespray's Avatar stevespray stevespray is offline
level 1
 
Join Date: Apr 2005
Posts: 10
Thanks: 0
Thanked 0 Times in 0 Posts
Rep Power: 0
stevespray is on a distinguished road
Thumbs up Re: JB's Freeeeee Neural Net Signals after closings

Batcheler….

fficeffice" />>>

Firstly, apologies, I certainly didn’t mean to sound too school teacher’ish and I accept that my post was perhaps lacking a required sensitivity. Yes, as you say, we’ve all been there before and therefore no one can ‘walk on water’.

>>

However, with all due respect, what I have pointed out is, in my opinion of course, correct. I’ll try and be rather more constructive with my comments which I hope will benefit the overall thread. So far it has been an excellent thread.

>>

Firstly, most of the more successful ‘mechanical’ FX trading systems which I have personally seen are ‘breakout’ systems. On such systems you would normally determine ‘a zone’ or ‘band’ between two levels and then play the market in the direction of any breakout from the predetermined ‘band’. The basis of such systems is to try and catch the bigger sharper moves.

>>

I feel that in this thread the focus has been more on trying to trade regularly and making a ‘regular’ profit. Unfortunately this is not normally the nature of the FX markets or, for the most part, the regular nature of any other markets. The basic premise of ‘Position Trading’ (which is what I personally feel your style is), is one where the majority of your money is made in a very low percentage of trades. The higher percentage of trades will end in a loss or, at best, a small winner. This as I reiterate is the nature of ForEx trading. With the loss rate being so high it is essential that losses are kept to a minimum. This is where most fall down. In order to take advantage of the BIG moves you have to have the funds to back them. This is why Capital Preservation is the key. Too many people end up in a situation where they either run out of capital or discipline before the big winner comes along. There is considerable psychology to be understood here – subconsciously you know that already; Your search for a ‘black-box’ mechanical system underscores the fact that you realise that you may not be suited to the ‘day trading’ approach for whatever reason.

>>

Moving to the system itself. Trying not to sound overly critical here, however I feel that points need to be made. If I can be honest I would suggest that you, and most of the participants of the thread, are looking for what experienced traders would call ‘The Holy Grail’ (THG). THG concept is an interesting one. It is actually a demand made of you through your conscious mind; the concept of having a system which is fully automatic appears to remove ‘emotion’ and pressures of ‘decision making’ from trading on a daily / weekly basis. This unfortunately is an area for ‘unlearning’. It is I’m afraid a very popular misconception. It is a misconception which makes many people a lot of money, unfortunately the people who make the money are the people who claim on websites, and in the financial press, that they can offer customers various systems which will return them X number of pips per month. Of course these ‘offers’ appeal to the uneducated because they appeal directly to the element of the mind which finds such offers attractive – an apparent trouble / work free method to make money – of course there is no such system, if there was then the people would not sell it to others, they would simply use it themselves and make all the money that a man could possibly need.

>>

Trying to be more positive now. The fact is that it is possible to build a ‘black box’ system which will work in most markets. Of course, as markets change and cycle, it will be necessary to re-evaluate the internal workings in an attempt to optimise trading without upsetting the positive expectation. This means that systems will have to demonstrate that they are ‘robust’ in nature – if optimisation is too finite then the advantage of that system can be lost in a very short space of time. This is where your problems start. By the way, I share you anger regarding ‘hindsight’ indicators – they are, in my opinion, the proverbial ‘ash try on a motor bike’.

>>

I have made a number of observations regarding what has been stated so far regarding the outlined system. The problem, in my opinion, is that the stops are far too loose. You will find that your losses will far outweigh your gains. You see, for most of the time, markets ‘vibrate’. They have what can best be described as a ‘natural frequency of oscillation’. This is why markets do not move in easy to trade straight lines. Technically, for the most part they trade about a central price which generally doesn’t exceed about 1.75 standard deviations of the ‘average true range’. This presents the potential system writer with a major problem. The key to trading any market is reacting to what the market is doing; we see a ‘set up’ and then the ‘set up’ gets triggered. Once the set up has triggered and placed us in our trade we are exposed to this potential market vibration. There is a high likelihood that, at some point, our trade will be pushed into loss. The problem is therefore determining when our signal to trade has been invalidated. This ‘invalidity’ can be established in two ways. Firstly, we recognise that a point is reached where statistically we can acknowledge that our trade was a bad one which now has a much lower chance of coming good and, secondly, a point is reached where an acceptable loss might be turning into an unacceptable loss (more commonly called a money management stop).

This is why most FX systems have such a high level of failed trades, in order for the overall system to produce a profit the high percentage of failed trades must be kept to a minimum – this is why these systems always seem to evolve in the manner which they do, they are adapting to their natural environment. The system outlined on this thread was one which kept you ‘in the market’ for an extremely high percentage of the time – this presents further problems. You are exposing yourself to potential randomness which you can not determine. This is not so much of a problem if the trade is nicely in profit, you are purely in a situation where you are speculating with the market’s money, this situation is altogether different if your trade is near breakeven or in loss. In this latter situation market vibration can push your trade further underwater even if ultimately your signal is going to turn out to be a good one. These facts further evolve the system required. It starts to be clear that a system is needed which detects a series of events which triggers a trade which will either be successful or fail in a fairly short space of time. Opening a trade is akin to the flying of a large aircraft – the most dangerous part is the ‘take off’ and those first early moments in the air; the aircraft has little speed and little height and therefore the pilot has very little scope for error should something go wrong – Human nature actually recognises this fact and hence the pleasant calmness which descends across the aircraft cabin a few minutes after you’re airborne. Likewise the same requirement may be useful to our potential system – because of our precarious nature at the opening of the trade we need to be more alert in case the trade is not conforming to what we thought would happen. There is therefore a requirement for system which implements a tight stop. My own extensive studies have revealed that stop of around 30 – 40 pips (assuming EUR/USD or GBP/USD) is the optimum stop which is required on a system which might fire a signal on a daily basis.

As system which can fire signals on a daily basis which has stops over well over 100 pips simply will not work – that is mathematical fact.

>>

You will find a good example of a ‘system’ on this page. I have no connection to this site by the way, I just find that the examples laid out are excellent.

>>

http://www.online-futurestrading.com/

>>

Click the link for ‘Trading Systems’ in the top left and go to ‘introduction’. From there you can walk through the various ways which you can develop your system.

>>

Hopefully, having read what is on that site, I can better make my second and arguably more important point. Money Management is actually the key to any trading system. Most people miss this fact altogether. All systems will have runs of winners and losers and this has to be taken into account when calculating how much ‘risk’ is going to be taken on each trade. The main reason that people lose money trading is because they suffer what is known as ‘irrecoverable drawdown’ – in other words a run of losing trades, which is statistically bound to happen, will cause their account size to drop by a sizable double digit percentage. Look again at that site and see how small the ‘edge’ that the actually trading system actually offers. It is something like 6 or so points on the Dow Futures – in others words, at face value, it looks a fairly unattractive ‘edge’. See however what happens when that edge is attached to a carefully designed Money Management system. Look how, over a longer period of time, the stakes and the account balance steadily grows. Suddenly, as I hope you would agree, the system looks far more impressive.

>>

I hope that the examples which I have given will help you in the right direction and kill off some of the false prophets which folk maybe following. All that is really required is a system which gives you 3 or 4 percentage points of edge over the ‘random walk’. From this, with good Money Management, a fantastic profit can be made. Unfortunately people do not follow this advice; instead they continue to search for ‘a holy grail’ which they simply can not find due to mathematic impossibilities.

>>

Best of luck,

Steve.
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!Spurl this Post!Reddit! Wong this Post!Twit this!
Reply With Quote
  #156 (permalink)  
Old 04-25-05, 04:41 PM
Batcheler's Avatar Batcheler Batcheler is offline
level 3
 
Join Date: Jun 2004
Age: 60
Posts: 327
Thanks: 0
Thanked 0 Times in 0 Posts
Rep Power: 9
Batcheler is on a distinguished road
eNNya-III and......uhh-hum....STEVE!

Steve,

Could you please remove your too large post. I barely glanced at such a thing, and it's so much yader-yader that I'm already familiar with. It is ENTURBULATING to the group, in intelligencia manner this is your intent. I am an engineer type, I know what I look for and am doing. I've already done few months of tests, particularly with eNNya-I. It is sound engineering I work towards, not pie-in-the-sky Grail stuff you accuse me of! So please remove or I remove my thread.

I now have eNNya-III ready to show folks, but you are stalling it with longiish giberish and enturbulating s@&t, so please, go to that post, and hit the delete or edit to get rid of most of it.

I was JUST now having a terrific session with tradecision using common ordinary TAs, and I have good results. Neural Nets are REAL learning artifical intelligence softwares, or do you not know? Study it on the web and you will find out.

You are new to this moneytec community as I found from some helpers research, so you are already giving yourself a bad name by doing this.

Remove it or I remove this thread and have my group that I have a list of go to a private relatively unknown quiet forum.

Like any other common sense guy on this thread, they try to ask Qs or make points or validate, etc, not jump on the scene with giant boots stepping on us all.

ATTACHED, anyway, is my HOLY GRAIL BUDDY!!!! LOL eNNya-III, no thanks to YOU! I spent the nite in contemplation, then 9am to 1pm workin' TD to the bone 'til I got it! Hey, 200-300 is NOoooo BIG DEAL, it just common sense FAIRLY consistent monthly gains, anyone with a pea brain can look at a chart and see "ok, we got swings this way, and swing that way....total for this or that month is..... if we could take advantage of....is.....xxx" Its CALLED "swing trading"!! THat's on web TOO! you should see how much pips some guys get with faster periodicities! This is daily bars.

Ah geeze, we sure didn't need this guy.

anyway folks, attached is chart + trades for walk-forward of eNNya-III, the TA bin is with just couple of TAs, SMA(3) and my confidential find that I put in a Velocity mode and Boolean IF mode to yield plus and minus signals SIMILAR to the AutoTrend Signals that helped eNNya-II so much........that's right, her work 'n now death had given us a clue of what is desirable, so that's why I'm of NO SLEEP now, contemplation throughout the nite and workin' it this morning with no sleep. So getting your wash job upon wanting to show folks something happy is not a happy thing as you just HAD to be Mr_Bowl_em_over_wash_job. Thanks a lot!

So beat it bud, go bother those other threads like "5 majors" and their 900 pip losses, we here have folks who did the demo tests for couple weeks and see its all real, not to mention my couple months tests before here on forum!

Your post is most disturbing to us all, delete it or I report you to moneytec webmaster.

You probably hadn't even read my Intro, have you?
 
jbs-freeeeee-neural-net-signals-after-closings-td_ennyaiii_10105jpg-jpg  
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!Spurl this Post!Reddit! Wong this Post!Twit this!
Reply With Quote
  #157 (permalink)  
Old 04-25-05, 04:59 PM
DKing's Avatar DKing DKing is offline
level 1
 
Join Date: Apr 2005
Posts: 10
Thanks: 0
Thanked 0 Times in 0 Posts
Rep Power: 0
DKing is on a distinguished road
Re: JB's Freeeeee Neural Net Signals after closings

I agree that there is no "Holy Grail" of trading however there is an ability of getting the computer to sort of do it for you. The key to doing this in my opinion is to have the computer double check results of its predictions, with the ability to change the trading system so that it can determine the best trade or set of trades at a given time within its ability (programming/ information avaliable to it) to do so. This would not be perfect but I believe you could get a good probability out of it, which is what it seems every trader deals with anyways. The probability that a given trade will have a positive outcome for them.
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!Spurl this Post!Reddit! Wong this Post!Twit this!
Reply With Quote
  #158 (permalink)  
Old 04-25-05, 05:04 PM
Batcheler's Avatar Batcheler Batcheler is offline
level 3
 
Join Date: Jun 2004
Age: 60
Posts: 327
Thanks: 0
Thanked 0 Times in 0 Posts
Rep Power: 9
Batcheler is on a distinguished road
Sorry folks, I'm tired, so blew my cork, there's no need for such characters

Sorry folks, I'm tired, so blew my cork, there's no need for such characters. Can you blame me, know I promised someone to keep cool & ignore, but midnite-oil plus this!? .......hopefully I'm excused.

If I were making grail claims woulld we have had results like we witnessed live for couple weeks? Tell this clown, PLEASE! If I be of no good to this group and community, speak up and I remove this thread! (and return donations!) Buuuut, up 'til this "G-U-Y" (!), NO one has behaved or been negative, most eveyone has seen and demo'd their way to a reality of what we're workin' on!

And this software doesn't lie, it gives real Sigs based on ones ability to handle it. Sure some wee-bit difficulty arises like it can for anyone, like a study function accidentally construed as a time series function. Folks here witness these things, along with money-management knowledge it's doable reality.

Geeze, I was starting this post to say something else and got into it again, sorry folks. Geee, I actually do forget what I was going to say, whew!

Anyway, the USDJPY looks ok too on eNNya-III.

I need to recover sleep-wise today & tonite. Hopefully we can crank-up eNNya-III in walk-forward in next couple of days. Ohhhhhhh, I know what I wanted to say! THe discovered TA function that acts like a sort of real time trend-'n-elbow-indicator is just that, real-time not some chart study, I double-checked 'cause we don't want to be fooled this time, and I've written Amit at Tradecision to triple-check that this time-series function really is time-series. We should here from him today or tomorrow on that.

bye for now. ZZZzzzzzzzzz
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!Spurl this Post!Reddit! Wong this Post!Twit this!
Reply With Quote
Sponsored Links

  #159 (permalink)  
Old 04-25-05, 05:11 PM
Batcheler's Avatar Batcheler Batcheler is offline
level 3
 
Join Date: Jun 2004
Age: 60
Posts: 327
Thanks: 0
Thanked 0 Times in 0 Posts
Rep Power: 9
Batcheler is on a distinguished road
thx DKing, but.....

Quote:
Originally Posted by DKing
I agree that there is no "Holy Grail" of trading however there is an ability of getting the computer to sort of do it for you. The key to doing this in my opinion is to have the computer double check results of its predictions, with the ability to change the trading system so that it can determine the best trade or set of trades at a given time within its ability (programming/ information avaliable to it) to do so. This would not be perfect but I believe you could get a good probability out of it, which is what it seems every trader deals with anyways. The probability that a given trade will have a positive outcome for them.
thx DKing, but.....if it did it for me and it did it for the group here (we both buy/sell pretty much the same time in demo mode), then its the real thing. What these walk-forward charts and "trades" inserts show is pretty much how myself as well as group will get, plus or minus so many wee bit pips. While it might seem deceiving, if you have it represent reality of the trades based on closes, then that's what you get. THe chart above w/ inset is just that, real walk-forwards just as if 1/2/05-to-now bars came to eNNya-III in random real life daily bar circumstance......so yea, its simulating reality. If one models reality then depending on how well you model it, that's what you get in reality. Engineers do that with lots of things, stress on bridges, heating on car engines, etc, etc.
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!Spurl this Post!Reddit! Wong this Post!Twit this!
Reply With Quote
  #160 (permalink)  
Old 04-25-05, 06:36 PM
forex777's Avatar forex777 forex777 is offline
level 3
 
Join Date: Jul 2004
Posts: 258
Thanks: 0
Thanked 0 Times in 0 Posts
Rep Power: 8
forex777 is on a distinguished road
Re: JB's Freeeeee Neural Net Signals after closings

Stevespray,

Thank you. Go away.

We understand these things. We, as a collective whole here, are walking within a cadence of "something different". Time will tell us about this effort.

Jerry is working like hell to make a little bit of history here.

But you can get step down off your crate and have a sandwhich.

Go to another thread. Right now, we don't need Mr. Know it all coming in here telling us how to manage the store. If we need a consultant, we'll send up the Bat Signal.

But, it's rude to come in here and tell us why we can't succeed. Please go count all the incredible sums of money you've made from your own trading. Clearly, you must be a billionaire by now.

I understand you must be stepping down to our level, purely for the purpose of Humanitarian reasons, us poor sods. If you really want to help, here is the toll free number to the Red Cross: 800 435-7669

You can concentrate your energies here. Or just go to www.redcross.com

Your incredible knowledge and understanding of the markets, and life itself, will ge greatly appreciated there.

Have a productive day..

Last edited by forex777; 04-25-05 at 06:39 PM.
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!Spurl this Post!Reddit! Wong this Post!Twit this!
Reply With Quote
  #161 (permalink)  
Old 04-25-05, 07:25 PM
stevespray's Avatar stevespray stevespray is offline
level 1
 
Join Date: Apr 2005
Posts: 10
Thanks: 0
Thanked 0 Times in 0 Posts
Rep Power: 0
stevespray is on a distinguished road
Re: JB's Freeeeee Neural Net Signals after closings

Ah DKing….Suddenly a man appears speaking sense who has at least seems to listen to reason. You are correct in your reasoning – a system which can check itself in an attempt to improve itself. There are such systems about which use what is known as ‘optimisation’ to improve potential trades, however, there is potential downside. As I mentioned before, ‘robustness’ must be a key attribute of the system. This means that ‘optimisation’ is only ‘fine tuning’ – it must not dramatically adjust too many key settings of the system otherwise any ‘edge’ is lost. It’s harder to explain what I mean – think of an ‘edge’ like a casino thinks of their ‘house edge’, it’s a fairly small percentage. Let’s look at single zero roulette as an example – the houses edge is 2.07% (1 over 37), see how small the ‘edge’ really is for these guys? Yet they manage to build and pay for ffice:smarttags" />lace w:st="on">Las Vegaslace>! Basically the casino’s edge doesn’t change and therefore doesn’t need to evolve (ie be ‘optimised’) – they just rely on the law of large numbers. Even if someone has a winning night the casino knows that enough people will lose to cover it, winning customers are very much part of the business just like losing trades are for the FX speculator. The casino knows that if $1m gets flung across their roulette tables they will pocket about $20,700 on average. In that last sentence the word ‘average’ is the key word. This is because it falls under the law of large numbers. This is where a potential system can come unstuck if it is not ‘robust’ enough. Over ‘Optimisation’ is what is basically know in the trade as ‘curve fitting’ – this is how these online advisories advertise such large returns – “We’ve pulled in 1200 pips per month on average for the last 15 months” – The fact is that they haven’t really done that, what they are advertising is a system which has be optimised so intensely that it is not ‘robust’ in any way. Their claims are purely based on back test results – the actual trades were never placed – read the small print and it is explained in there. Basically its ‘edge’ is a ‘freak’ based on how the markets happened to have traded for the last 8 months. I know plenty of people who have been sucked into such scams because they are unaware of what is actually required to ensure long term success. Poor folk. They see the website advertising the past winning months….1243, 1109, 876, 987, 201, 1345, 934, 1206…so they sign up. Four months down the line and they’re losing 700 pips per months – why? Because the system was over optimised, it may have had an ‘edge’ but that ‘edge’ is going to take time to work its magic, remember it’s a law of ‘big numbers’, that means that the more you trade it the more likely you are to get the results which your back testing has revealed. For example, if you tossed a coin twice then there are odds of 1 in 4 that you might throw two heads (in other words it can quite possibly happen), however, throw it 1000 times and you’ll end up with a far more even set of results and the results become somewhere nearer 50/50 (ie they didn’t all come up heads even though each was an exclusive event). Hope that folk can see my point. If a system is continually optimising itself then there is a risk that its ‘edge’ never gets a chance to show itself, as soon as a few losing trades are observed the system is reconfigured so the recent ‘bad trades’ are ironed out. It could well be that the losing trades fell within the constraints of what had been determined by the initial back test but the pressure is on to produce instant profits and not loses.

fficeffice" />>>

The problem is, when a system is designed, the natural inclination is to look for extremes of performance. Unfortunately, with extremes comes a lack of ‘robustness’. It’s the nature of the territory – ever noticed how nice champagne glass always seem so fragile!

>>

DKing, you use the word ‘prediction’…tut tut…You can’t do that. We do not ‘predict’ things, we are slaves to probabilities and the laws of mathematics. What we look to trade are ‘probability events’, a set up where we know that the is a slightly higher chance of A happening than B happening – I hope that you can understand that. Just the same as the casino with their roulette table.

>>

Forex777….

>>

Sorry for my rudeness. As I mentioned before, I did not wish to cause offence. I am certainly not ‘Mr Know it all’, I am merely imparting knowledge which I have discovered my self over many years of trading. For what it is worth, my main experiences have been in the American Futures markets rather than ForEx. I have however spent many hours / days / weeks / months studying all kinds of markets, FX included.

>>

I’m obviously missing a song here….you’re looking for ‘something different’…I guess I’ve kind of missed the point then.

>>

Batcheler….

>>

It certainly wasn’t my intention to ‘hijack’ the thread in anyway. What I have posted is my opinion and only my opinion. People have the choice to decide if any of the rambling that any of us post is relevant to them or what they chose to do in life. I don’t understand why you feel threatened by the posts – all they impart is information, just the same as your posts. By the way, I can not open that private message, apparently I need to have made 25 posts in order to access the private messaging service.

Obviously if you feel that my post is offensive then feel free to report it as such.

>>

As to your post. You mention Neural Nets and AI…… “Garbage in – Garbage out” as someone famous once said (when talking about computers). Unfortunately it is the truth – A computer can only work with the information that it is provided with, there is no facility for it you use its ‘imagination’ as it has none. Computers, when they are working correctly, do exactly as they are told, nothing more, nothing less. On that basis it is impossible for you to expect the computer to develop a strategy any different to the one which you program into it. Sure you can set it running testing ranges of data which it enters into a defined number of variables of your choosing but that isn’t a NN it is pure ‘optimisation’ – basically finding a set of numbers for your set of variables which gives you the best results for you chosen back test period. Could you share with the group the make up of your ‘black box’ model?

>>

Steve.
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!Spurl this Post!Reddit! Wong this Post!Twit this!
Reply With Quote
  #162 (permalink)  
Old 04-25-05, 07:48 PM
Batcheler's Avatar Batcheler Batcheler is offline
level 3
 
Join Date: Jun 2004
Age: 60
Posts: 327
Thanks: 0
Thanked 0 Times in 0 Posts
Rep Power: 9
Batcheler is on a distinguished road
thank you mr steve

thank you mr steve. anything else you would like to say? appreciate your inputs.
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!Spurl this Post!Reddit! Wong this Post!Twit this!
Reply With Quote
  #163 (permalink)  
Old 04-25-05, 08:05 PM
bigcat's Avatar bigcat bigcat is offline
level 1
 
Join Date: Apr 2005
Posts: 21
Thanks: 0
Thanked 0 Times in 0 Posts
Rep Power: 0
bigcat is on a distinguished road
Re: eNNya-III and......uhh-hum....STEVE!

Jerry,

In eNNya-III I can see that some of the signals now occur slightly after the peaks, on the bit of the EUR chart you posted, which seems more realistic. Results there look promising. It also looks as if eNNya-III resolved this last run down in the EUR, and is now again in a short from about 1.3048.

Are you thinking of returning to the daily calls for EUR and JPY, based on eNNya-III ?

Great work. Also, nice to see you keep an even temper with regard to Steve's posts. I imagine that if people are interested in them, they will reply to them. But if they see them as off on a tangent, nothing will happen with them. Steve's a smart guy and probably not going to repeat his points, they were well enough said.

What I am interested in is what YOU have to say. And particularly, what you have to say in the near future about eNNya-III !

Cheers, Bigcat
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!Spurl this Post!Reddit! Wong this Post!Twit this!
Reply With Quote
  #164 (permalink)  
Old 04-25-05, 08:14 PM
bigtex's Avatar bigtex bigtex is offline
level 1
 
Join Date: Mar 2005
Posts: 5
Thanks: 0
Thanked 0 Times in 0 Posts
Rep Power: 0
bigtex is on a distinguished road
Thumbs up Re: JB's Freeeeee Neural Net Signals after closings

Hear, hear! Well said BC!
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!Spurl this Post!Reddit! Wong this Post!Twit this!
Reply With Quote
  #165 (permalink)  
Old 04-25-05, 08:28 PM
Batcheler's Avatar Batcheler Batcheler is offline
level 3
 
Join Date: Jun 2004
Age: 60
Posts: 327
Thanks: 0
Thanked 0 Times in 0 Posts
Rep Power: 9
Batcheler is on a distinguished road
tad explanation before eats 'n rest tonite

tad explanation before eats 'n rest tonite:

basically, after finishing my job this morning I had built-up steam to do what I clearly and vividly recall in my younger days as an electronic engineer, that of a "Peak detector". I was amazed that there was one TA function such that when Velocity is applied to it you get a curve that crosses zero whenever a peak is reached. I then put that expression into a Boolean logic formula that yielded a plus or minus depending on whether a higher or lower peak is going on. Thing is with a peak-detector, when it swings going other direction that means a trend, tiny or big, is somewhat or fully (after time from that event) going on.

We still have that "kluge idea" to use eNNya-II: just tolerate getting cut-out of trade before peak, with back-up by eNNya-I to give us signals in meantime, but now this "eNNya-III" discovery of a TA combo for a peak detector is coming along fine, so I'm going to track its events for awhile. Its so completely robust due to giving eNNya strong trend/peak/valley Sigs that I feel that just doing even a short while will show everything is ok, although its almost silly to do real-real time tests because just picking different walk-forward dates already simulates reality enough.

(you know, between you and me folks, I always knew I had some sort of disdain for intelligencia types, well, above muckracker proves it out, they just want to bedazzle you with obfuscation and filibustering, poor souls, you know, like "get a life" applies, eh gang? LOL if they only knew how ragged-out we feel when they act such......unfortunately some may be on the evil end a tad too much so enjoy doing it....sad.)

Hey, I got an engineer streak in me, so when I see opportunity to give something a whirl I take a look. If its a strain for some dragging thru some of this......granted we all thought original eNNya was it.......I had no intentions of doing engineering while on a forum, just wanted to share what I had......so sorry if its a bit of a drag for some, thx for bearing with all this.


Thanks for being here with me.

OK good folks on my list, good nite!

Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!Spurl this Post!Reddit! Wong this Post!Twit this!
Reply With Quote


Reply

Bookmarks

Thread Tools
Display Modes

Posting Rules



All times are GMT -4. The time now is 11:29 AM.



vB Enterprise Translator (vBET 2.3.10) made by NLP-er


no new posts