/biz/ - Business & Finance

>>7843267
So green was calculted befor blue happened?

>>7843325
green was calculated without knowing the blue line

>>7843267
the absolute state of 6kers

>>7843267
What parameters are you using?

>>7843267

This looks like a LARP chart later fabricated.

Show proof.

>>7843348
proof of what?

>>7843347
it's a lstm network after 300 epochs of training

i'll try to find some hourly data so we can test it through the day and see

i'm training it on hourly data, let's see how good/bad it will nail the next 14 hours

>>7843267
thank you for your efforts anon, looking forward to seeing more posts

>>7843537
i will if the prediction nails atleast the next 7 days trend.

btw, posting hourly forecast in a moment

>>7843267
do you have a closer view by chance? A 1, 2, or 4 hour chart?

>>7843558
Ah perfect tyy

this is what i get on the hourly chart

>>7843608
predictions are way cooler if you are risking real money, kek

>>7843608
Well I'll be

>>7843608
already creating a recursive mirror of that chart

Pic related
OP

>>7843652
>1m
kek

i'll close my long in exactly 6 hours from now and short there, we'll see how it'll go

>>7843652

something is wrong with your computer screen there's a bunch of lines on it

>>7843666
actually the training data was way longer, like 1 month, including the dip to 6k

>>7843668
LOL
BLOCKLOOP

>What is overfitting?

>>7843668
nah you just panic bought at 11,750 faggot stfu

>>7843685
those are just support/resistance lines lol. need to clear it soon

Whats the difference between original and training data?

>>7843704
it's pretty generous with the closing prices, doesn't look like it's overfitting that much. as long as it nails the uptrend/downtrend it's enough to make money

>>7843759
original is the real btc price i gave in input. then the network trained itself by using the data overlapping the orange line.
after that i asked to predict the green one without knowing the underlining blue one and you see the results (though it used the blue one at the end to verify how accurate the prediction was).
red line is completely made up from the latest data

>>7843759
im guessing it is his model performance on training data.
So basically past performance on data that is included in the model's memory.

>>7843796
exactly, i think i still need to tweak some variables but overall it looks good enough

>>7843792
Is this a program or a manually configured/coded dataset API type deal?

>>7843792
You dipshit.

You need Train/Test/Validation if you want to put a model like this in production, otherwise you are just manually overfitting on your test set.

>>7843815
python and tensorflow, but the model itself is manually coded

>>7843828
i did train/test/validate, what's your point?

also what production? if it works i'll use it to make monopoly money and that's it

>>7843337
If this is legit he wouldn't have sold at 6k and sold at the top like a normal person

>>7843847
Have you tried convolution nets?

>>7843668
kek

>>7843847
What you probably did is Train/Test-split

You've basically overfitted on your test-set by using it for parameter tuning.

The problem with this prediction and most predictions in general is that if an influential amount of people from the outside act upon them, the outcome of the prediction changes dramatically

The market is like a bunch of vectors in a pot, and outside forces just keep throwing vectors in.

>>7843608
Have another? I was a cunthair away from having my short liquidated and now this is happening. We gonna keep dumping?

Well, yeah, theres a bear flag forming (1hr) and we hit resistance + below the cloud on the daily (perfectly) ...dno if your shit is legit or not, but we're going lower for sure when the daily MACD flips red

>>7843902
I think it will bottom out in the 9000s personally

>>7843267
Are you the same anon that was posting a few weeks(months?) ago as AI anon?

>>7843902
At least you weren't this poor sap 3 hours ago

>>7844033
God damn

>>7843867
i'm doing it as a sideproject, a 1-d convolution net, with convolution layers being like "timeframes" but it's harder to setup properly so it'll take some time

>>7844025
no

>>7843848
i didn't sell at 6k but i've finished the network yesterday

>>7843902
i've updated it with the last hour and it's still more or less the same, though i wouldn't rely that much on the hourly data

>>7843902
>>7844022

There's no way in hell it'll go below 9k. That line of support is simply far too strong. It'll probably go sideways for a good few weeks once it dips below 10k.

>>7843267
so epic anon. gonna watch your threads more
what 'language' or platform do you write this in

linuxfag here

the best thing is you can just be a very elaborate troll

>>7843267
so...how low are we going here? 4k? by first week of March?

>>7844110
my thoughts as well lmao

>>7844134
4k is very, very far out of sight at this point. more like 5k absolute bottom. Highly likely bullish continuation imo

It won't even break 10k before 27k in May

>>7844080
stfu moron

>>7844105
python and tensorflow

>>7844134
>>7844240
i actually agree we won't see 4k.

actually i doubt we'll even see 6k. the actual predicted prices are pretty wonky, but 99% it'll drop sub 8k during the next 7 days

>>7844421

Update >>7843608

It's been more than an hour. Let's see if it was correct.

>>7844492
it's still similar. peak in 4-5 hours and then dump

>>7844511

Post the updated picture.

>>7844421
yeah post up OP

>>7843267
whats it trying to predict? one day in advanced?

>>7843608
for the record, the heigh of the red line in hourly predictions is kinda arbitrary, because my chart was set up for daily values and i can't be bothered fixing it right now

>>7844535
>>7844570
here's the updated one, look like we'll see 11150+ one last time before going down

>>7844614
2 week, but i expect it only to get the main trend of the next week desu

These posts are 100% fud. Trying to get weak hands to short and get ass cleaned. If you think Btc is going back sub $10k before June ish then you’re retarded and high

>>7844691
but i'm long now... >>7843630

You all realise this is a larp because this idiot sold at 6k right, why on earth would someone make a machine learning program and post it on this shitty site

>>7844691

The absolute state of forever bulls.

>>7844731

You really shouldn't have all inned with your rent money. Americans are going to sell off as soon as they wake.

Bear & bull banter on here is adorable, you guys should just fuck

>>7844772
Topkek go back to being cucked eurofag

>>7843267

How are you importing the data for training? Are you just using price as features?

It would be interesting to test different algorithms with additional features like RSI and other metrics traders use.

>>7844750
Btc in downtrend for months.
Uptrend for a couple weeks

AHAHAHA U MEAN IT MIGHT KEEP GOING UP ? FOREVER BULL UNCNA MAKE MONEY WHEN IT DIPS YOU KNOW ITS ABOUT TOME IT CRASHES AND U KNOW IT.


You sold at 6k and it shows, made more money shorting btc than going, I want it to break support because then I can put a massive short with high leverage and kill it in less than 24 hours, going long is riskier most of the time, but I’m not as deluded as you are weak hands faggot.
Anyone calling for btc crash when it’s above 2 important supports and just had tiny pullback from a long bull run is either coping with the fact he sold low and fudding or straight up brainlet.

>>7844691
Did you miss the -1100 usd drop earlier over the span of 100 minutes or are you just braindead?

>>7844850
Absolutely btfo

>>7844839
yeah, only closes. i grab csv from yahoo finance or coindesk

>>7844839
>It would be interesting to test different algorithms with additional features like RSI and other metrics traders use.
while predicting the real price is just a meme, people predicted moving averages since the 80s
. you can do a lot really

>>7844864
Just bot selling after EMA cross which hits some stop-losses and margin longs.
Now we go up again.

>>7843267
Don't let these fags FUD you OP. I'm sure some hardcore investment firms have been using this tech with crypto and stocks to great effect.

>>7844033
wait is that $1.6 mil or 1.6 mil BTC? brainlet here

>>7844937

Sub 11k again

>>7844963
$1.6 mil

Monitoring this thread, thanks based op

>>7844902
I for one support your endeavors OP these dumb fucks are pretty pathetic trolls

>>7844963
USD lol

>>7844683
thats pretty neat, im trying out a CNN to convolute the time spans instead of features.
whats your ROI?

>>7845055
>whats your ROI?
we'll see it in a week, lol

actually i think that predicting prices is a meme. my next project consists in using deep q learning to train an agent to trade. that's actually doable and provably working

good job anon, what kind of network are you building for this? do you have it in git?

>>7844938
>comparing investment firm ML models using advanced stochastics and tuned indicators with OP's shitty lstm/CNN on BTC price.

Here is a tip,basic regression techniques are more effective than any machine learning technique when your dimensionality is low.

How about the OP trains his model up to January with random sampled test data and shows us a blind February prediction.

>>7844850

The salt is delicious.

>>7844850
>going long is riskier most of the time
the absolute state of /biz/

>>7845094
>provably working
have you tried it? i agree some sort of reinforcement learning needs to be tacked on, but for me i cant find the right NN structure thats workable with crypto or stock data. LSTMs didnt work for me, hence CNN. maybe a combo of both.

>>7845172
for provably working i meant that since there exist traders who can consistently make money, then you can train an agent doing the same or even better. i also know some people who did similar stuff for their thesis and now they work as analysts in jp morgan or some banks

>>7845246
What line of education would this be considered?

>>7845284
computer science/financial analyst or something like that

>>7845246
How old are you? When did you start studying this stuff? Just curious and well done anon. keep up the good work

>>7844683
Holding onto my short still.. hoping for tendies and not a liquidation email. Fingers crossed!

>>7844850
>Tiny pullback
>1000$ crash in 3hr
>Needed 2 days to get up from 10600 to 11700

This is how you know we are in bear market.

>>7845430
I think you've made the better decision thus far, rising wedge pattern atm

I opened a long at 10800, fuck now I want to close it and go short instead... Hold me /biz/

updated with the latest hour

the red line should be high, it's a charting problem, for actual targets use ta.

it's say it's still consistent enough, predicting another small rally and then dumping.

>>7845430
i'm still keeping my long open for another 3-4 hours and then i'll short, but it doesn't look like it won't touch 11500

>>7845412
almost 24, i'm getting a master degree in cs next year.

i actually started studying programming in 2010 because i wanted to make a minecraft clone, lel

So far honoring both the rising wedge and bearish consolidation/pennant.. leaning toward shorting atm myself

Anon. Can you create a discord and share your based technicals with us? You seem to have a good handle on this tool.

>>7845580
Top and bottom diag lines are the rising wedge lines

>>7845520
That red is below the earlier lows? I shorted that spectacularly and then got greedy. Paid for it with a sleepless night and a 9-5 I gotta be at shortly. Yuck.

>>7845622
the red line should be higher and lefter, it's a charting problem

>>7845580
So many lines can barely tell what's what

Don't mistake this as FUD, just TA, but it's also still heavily overbought on the 2, 4, and 6 hour ichimoku cloud

>>7845647
I'm used to it lol sorry, zoomed in too much

>>7845580
do it faggot. short now before amerifats wake up and tomorrow china resumes activity. also friday is futures so yeah ... it's probably the last chance to short for a few months.

>>7845689
considering it more by the minute

>>7845721
wait 2-3 hours, we are pumping now

>>7845884
I'll take your word for it with the buy rush just now

>>7845721
Both longs and shorts are stagnant atm, it's better to wait for them to pull in either direction before doing anything.

OP make a discord channel

>>7845905
You know if it does cut one way or the other, down is the likely direction. Another shirt lived pop on the 5 minute

OP what should I do at uni if I want to do this? I'm a math guy, is that close enough or CS only?

>>7845906
been thinking the same, if it breaks the wedge it'll be hard not to long

>>7846033
machine learning stuff is more math than cs desu. just be sure to do some basic programming course, learn python, and then focus on machine learning, which is 90% math anyway

>>7846081
I'm not the college type but I'll do some reading no doubt, appreciate ya

>>7846081
Is there a "machine learning" course at your uni specifically? I don't think we have it here

So learn python and I guess more applied math? thanks anon

400 btc sell on bfx, kek

OP, do you use any indicators?

>>7843267
using TF with Keras, OP ?

You could actually win a couple bucks with that, you know. Like you could create a website with an API that gathers data from your machine learning bot and feeds it into a nicely customizable graph or something.

>>7846169
yes, i actually have a machine learning exam. machine learning is a pretty extensive term that covers a lot of things, like classifications, data mining, predictions, agent training, etc.

>>7846247
rsi and macd, sometimes ichimoku memecloud

>>7846326
yes

>>7846334
damn smart anons here today

THIS RUBBISH ONLY WORKS IN A TRENDING MARKET DUH. All you are saying is it will go down.

Try calling the exact time and price of a pump or dump with a track record then I will have some respect. Unless you put your own money up and prove it OP your stats are meaningless

Yesterday Darktrip called the exact price level and time of the crash and put his/her own money up on massive leverage on bitmex.

Cool stuff OP, keep it up

>>7846405
You looking at the charts smartass? because the shit is plummeting, so stfu

>>7843267

Hey buddy!

I'm in the process of setting up an LSTM as well, may I ask, what did you feed in other than the sequence?

>>7845094
You should post around this time in a week so I can see the progress anon, pretty cool thread.

anyway, the hourly prediction will be shit until i fix the exact red line placements so fuck those for now.

i'm still confident that the red line on the daily is accurate for atleast 7 days, so i'm shorting now and leaving it open till march (use ta to place entries)

>>7846473
i only used the price sequence and i may add volume soon too, but i heard that if you get daily sentiment analysis on btc news articles you get pretty impressive results.

>>7846479
yeah, i will, even if it's completely wrong you'll be able to laugh at me in 1 week

>>7846508
Quantified sentiment analysis, holy shit lol that's great

>>7844902
Thanks for the tip anon. I work with machine learning algorithms and robotics for a living, so I've been interested to try ML based methods for prediction. I might try a binary buy/sell classifier.

>>7846613
training an agent who buys/sells stocks is a real possibility
http://prediction-machines.com/wp-content/uploads/2017/07/Python-Meetup-Presentation.pdf

>>7846433

saw this when you posted earlier today or yesterday, really fuckin' neato. i was saying 8k last time.
the algorithm is saying whales are greedy as fuck, wanting to crash it and then scooping up cheap for 6k again. but you can't fool people the same way twice, everybody will be waiting for 6k.
they will stop it at ~8.3k so all the plebs will be waiting, they don't get in and then have to fomo in at 10k.
so, are they smart or will they be greedy?

>>7846679
Yes, I've done some reinforcement learning in the past. The only thing I'm unfamiliar with is how to interface with an exchange.

>>7846679
hope you closed your long OP!!!

>>7846706
here's another prediction with a little different parameters, which is think is more realistic

i'm 99% sure we'll touch 8k, but other than that it's pure speculation. i'll keep updating the prediction daily to see how consistent it is.

also don't follow this predictions literally. if my network predicted 7 days of downtrend since yesterday, that mean you should actually wait for a trend reversal and short on a retrace.

machine learning aside, btc wil dump during this week anyway

>>7846749
yes, i went short right sub 11k on the retrace

>>7846730
that's the easiest part, you just make some api calls in php/python/js or whatever you like

>>7846698
my fucking sides

>>7846706
>so, are they smart or will they be greedy?
Actually nobody will be waiting at $6k. Without reaching a new ATH it should keep going down since there's no hope anymore.

>>7846679
>http://prediction-machines.com/wp-content/uploads/2017/07/Python-Meetup-Presentation.pdf
whats the first letter or your name?

>>7846836
it's not mine and i found that by random googling

>>7846836
>mfw you niggers want to ruin everything

>>7846854
hahahahahaha if only you knew

>>7846778
Op thanks, my short worked out well

>>7843267
Show me your short then faggot

>>7846873

>>7846873
>showing your position on /biz/ where bots are scalping to liquidate

>>7846903
>bots pushing price adove 12k to liquidate my 0.2btc
kek

>>7846873
go cry about your empty wallets and stop fucking with the only good thread on this site

>>7846928
classical /biz/ I remember back before the drop to 6k people where sharing TA and discussion BTC potential drop, those threads where completely shat on.
/biz/ is all about 5 Link threads, and the current flavour of the month shitcoin threads.

ok, my last serious contribution for this week:

the uptrend objectively reversed, so we downtrend now. also my meme chart prediceted this since yesterday, plus we touched the downtrend memeline on 1d chart and bounced down.

the only sane thing to do now is to go fiat/short on a retracement, or atleast don't keep your longs open.

according to my predictions, the bottom will be somewhere on 25-26, so i'll keep the short open till then. don't trust memes, nobody can predict the actual highs and lows.

use ta for targets/entries and exits (for example a 0.5-0.618 fib retracement is a pretty safe bet, with bottom being 8k-8.5, but it may go lower)

>>7847022
Thanks for making this thread and sharing your thoughts and charts. It's nice to see a proper thread every now and then.

>>7847022
Thanks anon. Quality thread in a sea of complete and utter shit.

>>7847061
>>7847086
np, i'll post more predictions next week if this one will be at least somehow accurate.

>>7847022
whats the structure of you ltsm? how many hidden layers/nodes in your network(s)?
also how many past outputs do you consider for each prediction and what inputs do you take?

>>7847022
>>7847061
>>7847086
it's been a pleasure anons, thanks for the info and thoughts. I'll be searching for a followup this time next week

>>7847022
>Thanks anon. Quality thread in a sea of complete and utter shit.
I agree with that

you're not coming back this week?

>>7844902
Only closes kek.

I have tried using LSTM for months now. The only somewhat accurate predictions I got when including much more data. Grab more data, perform a PCA and you will see why your model is just a glorified RNG.

Greetings from an experienced AI anon who does not share his models to create fud.

P..S. double check that you are not overfitting, because your model sure looks like you do.

>>7847211
Addition to all other anons: Do not trust a prediction you have not made yourself.

AI is often marketed as a magical tool, which it is not.

When you want to puke on your shoe after waking up in the morning, YOU BUY.

>>7847240
Also good advice.

Always good to blend sources & intuition to some degree tho

ITT : One anon suffering from The Dunning Kruger Effect and many luddites who think he is right because they don't know technology

protip kids : if it was this easy that one random NEET could do it and make money, wallstreet or VCs would have been doing it years ago

use your brains once in a while

>>7847307
>lstms literally invented in 2014
>muh wall street dinosaurs have had this technology for years

>>7847307
>The Dunning Kruger Effect
Aww anon learned a new term from his entry psychology class in 10th grade

you sound clueless and bored lol, if you actually watch any charts you'd see that this info has been accurate for several hours now. gtfo poorfag

>>7847307
5/10 troll

>>7847387
>Institutions employing PhD Mathematicians and cutting edge IT equipment (look up TCP offload engines) are behind the times! It's us NEETs that are the future

unwarranted self importance

>>7847393
It's right, until it's not. Even the Big Boys run risks with this (flash crash, positive feedback causing huge losses instantly, etc)

>>7847307
This anon gets it.

I do also show my models to my friends from time to time but I tell them that the prediction is most likely utter bullshit (which is very accurate). And my model is much more advanced than the one described by OP.

I even use NLP for sentiment analysis.

DO NOT TRUST ANYONE JUST BECAUSE HE KNOWS BASIC AI STUFF.

You're going to need something more advanced than a lstm to make money kek

>>7847415
>I even use NLP for sentiment analysis.
Nice, that's about the only thing i would put stock in frankly.

>>7847415
No shit sherlock it's just a discussion & an attempt to gauge wider trends. we're only fucking with hourly charts at all because it was a deciding moment in the continuation/reversal.

Fuck microtrading

>>7847387
>Thinking time series machine learning analysis started with LSTM's
either very young or actually just stupid.

I love how this entire thread has flown by and no one has mentioned how retarded OP is for not taking random samples for his test dataset and chooses just to use the past month. Its incredible someone with zero understanding on wtf hes doing has the balls to tripfag on /biz/ as "MachineLearningAnon "

here is some python which should yield similar results to your model OP.

>random.randint(0,100)/100

>>7847022
>>7847211
College dropout doing online courses in ML/ data sci. Am I a meme who is never going to be as decent at predicting stuff as graduates/ PhDs

>>7847601
Your skill does not depend on a degree. A degree makes getting hired a lot easier though.