GNY Range Report - Free AI tools for traders.

[John_Muck]

I agree with Cobgoblincave4 that we are hitting a bit of a bump in the road with the accuracy figures for XRP; currently sitting at 75.54% when forecasting price volatility. I am confident though that we will get it up into the 90's like many of the other tokens we cover.
Latest forecast and historical for XRP below as an example.

So how do you improve the accuracy? Do you feed new data or is there any tweak to the algorithm? I believe some coins are just wildly volatile since their fundamental changes quickly like XRP after the SEC case, so it is not surprising if your current model failed to stay within 90% like other coins. People should be aware that this 'prediction' can't correctly express any change in fundamental values if I get your explanation correctly.

Good question. First off it should be noted that such a undesirable dip in accuracy will not cause a kneejerk reaction from the team resulting in unnecessary fiddling with the model. So for example something like an undesirable outcome of an SEC court case can tank the price of a token out of nowhere and also catch the ML unawares thus driving its accuracy percentage down. We would simply leave the model to readjust its forecast to the "new normal".  If it did not adjust for a prolonged period of time then steps could be taken.

GNY maintains a data warehouse containing all available historical price information for the twelve supported tokens and twenty five of the most popular and predictive chart analysis trading indicators. Our Fabian model is trained and tuned using this historical data.

On the first of each month, Fabian automatically conducts a hyperparameter sweep in order to tune itself and ensure that the model is performing to its maximum capability and efficiency.

Compared to the previous machine learning model, Fabian utilises more data points and those data points that it uses have greater granularity. For example, the original model used 1 day candles and indicators whereas Fabian uses 6 hour candles and indicators. This allows Fabian to be more responsive to predicting the likelihood of short term volatility and relative value changes. We are continually adding to this curated pool of data points and we believe this can lead to more accuracy and further-looking forecasts.

[stonehedge]

Just to go into a bit more detail...

The current release of our LSTM model (Fabian) has a semi automated tuning process.  On the 1st of every month all 12 models runs x100 times, testing different hyperparameter values.

We then look at the model predictions for each of those 100 runs for each model to find the hyperparamaters that are most predictive in that moment.  This is currently a manual step, I can assure you that checking 1,200 model runs is a real drag!

Now that we've been testing for a few months I've started working on automation for that part of the process.  If testing goes well, the first fully automated tuning process will happen on 1st October.

As John_Muck says, the trick with managing ML models is to figure out a perfect balance of knowing when to intervene and when not to intervene.

No ML model can be predictive all of the time (not many people realise this, given how many "AI" companies out there claim to be perfect) but its also possible to over compensate and ruin a model by tuning too much. Overfitting is one problem that we have to avoid...when a model learns the training data too well, including its noise and outliers, but does not generalize well to new, unseen data. Essentially, an overfit model performs well on the data it was trained on but poorly on any new data.

[stonehedge]

Is there anybody on this thread who would be happy to give us 30 minutes of their time to take part in some structured user feedback testing for us?  

We're offering 5,000 GNY tokens in return although at this point we're limiting it to five people. The aim is to record a small number of experienced crypto traders navigating Range Report and attempting to complete some specified tasks so we can assess how well (or badly) our site is layed out.

We're going to be using userlytics.com to do this, its a pretty cool system.

Please DM if interested.

[John_Muck]

One way we are improving the accuracy of our forecast is with a trial of the incorporation of WOTC (wisdom of the crowd) data into our models. This begins with BTC.
We are currently running an airdrop/bounty program to help us access BTC price forecasts from members of the public.
We would love to get your predictions and reward you for your time.

https://www.gnyrr.com/airdrop/

Predict, Share, Win: Your Crypto Trifecta!
Unveil your crypto savvy in our thrilling Airdrop Bitcoin Price Prediction contest! Guess Bitcoin's price on Binance.com at 0000 UTC Friday October 27th, 2023.

Your prediction will be used to train the next generation of our predictive machine learning models, to help improve our accuracy.

Submit your forecast FOR FREE between 0001 UTC Monday October 23rd, 2023 and 2359 UTC Wednesday October 25th, 2023 to snag up to 300 $GNY BEP20 airdrops!

Each action you complete earns you 50 $GNY BEP20 tokens. Amp up your rewards with extra tokens up to 300 $GNY BEP20 by buzzing about the contest on social media and referring your friends!

Everyone who enters gets rewarded, and the closest prediction within 3% of BTC's closing price Friday October 27th 2023 grabs an additional $100 prize.

How sharp are your market predictions? Show us your skills, spread the word, and earn $GNY just by playing!

If you have any questions I am happy to answer them below

[John_Muck]

Some good predictions coming in to help our WOTC input to our data.
However the jump in BTC's price in the last 24 hours will have shook things up quite a bit  
Remember you have until 2359 UTC Wednesday October 25th, 2023 to snag up to 300 $GNY BEP20 airdrops by taking part and submitting your BTC price prediction.
The Link to enter is here: https://www.gnyrr.com/airdrop/

Please note: The GNY Airdrop is safe. However, we would always recommend that people use a new empty wallet for airdrops. Doing this will prevent you from losing money to scam airdrops from less trustworthy projects. 🛡️Stay safe, and the best of luck with your entry.

[John_Muck]

Just to warn you to get your entry in to snag up to 300 $GNY BEP20 airdrops! as this is closing at 2359 UTC Wednesday October 25th, 2023

https://www.gnyrr.com/airdrop/

Predict, Share, Win: Your Crypto Trifecta!
Unveil your crypto savvy in our thrilling Airdrop Bitcoin Price Prediction contest! Guess Bitcoin's price on Binance.com at 0000 UTC Friday October 27th, 2023.

Your prediction will be used to train the next generation of our predictive machine learning models, to help improve our accuracy.

Submit your forecast FOR FREE between 0001 UTC Monday October 23rd, 2023 and 2359 UTC Wednesday October 25th, 2023 to snag up to 300 $GNY BEP20 airdrops!

Each action you complete earns you 50 $GNY BEP20 tokens. Amp up your rewards with extra tokens up to 300 $GNY BEP20 by buzzing about the contest on social media and referring your friends!

Everyone who enters gets rewarded, and the closest prediction within 3% of BTC's closing price Friday October 27th 2023 grabs an additional $100 prize.

How sharp are your market predictions? Show us your skills, spread the word, and earn $GNY just by playing!

If you have any questions I am happy to answer them below