ITB Predictions available in Turkish Exchange Ege Money

Subscribed exchange users can access ITB’s one-hour directional price predictions

Ege Money is a leading exchange in Turkey and the Aegean Sea region. With a unique variety of features such as portfolio management, vault protection, recurring buys and mobile compatibility, Ege Money offers their users a 360 degree experience when it comes to the trading of crypto assets.

Starting today, Ege Money users will be able to access IntoTheBlock (ITB) analytics as well as ITB directional price predictions for those assets covered by the ITB Platform. When users navigate to each crypto asset’s page, they will see two tabs. The Token Summary tab displays, on the top part, several of IntoTheBlock’s indicators based on blockchain, price, social and other datasets on top of which statistical and machine learning models have been applied. The bottom contains momentum signals based on IntoTheBlock’s underlying indicators.

The “Price Predictions” tab shows IntoTheBlock’s powered one-hour directional price predictions for the next hour. The dashboard displays not only the most current prediction but also the last four predictions. Additionally, in the spirit of transparency, the accuracy for the past 7 days and last 24 hours is also provided.

What is ITB Prediction Services?

ITB Prediction Services is a platform that uses deep learning models to detect patterns and forecast directional movements in crypto assets. The platform was designed to support all sorts of predictive experiments many of which are very specific to a given crypto asset. For instance, you can find models in ITB Prediction Services based on order book datasets which can be applicable to different crypto assets while other models could be incredibly tailored to a specific crypto protocol.

ITB Predictions Services builds on the large datasets aggregated by the core ITB platform. Today, ITB includes over 50TB of data related to crypto assets and incorporates new data every day. Most of the data can be categorized in one of the following groups:

· Blockchain Datasets: Transactions, blocks, address, etc.

· Spot Order Book Datasets: Trades, orders, etc.

· Derivative Datasets: Futures contracts, perpetual swaps contracts, trades, etc.

· Alternative Datasets: Twitter, Telegram, Google searches, GitHub, etc.

You can think about ITB Predictive Services in three main blocks. The first block includes a curated group of datasets used to train the different predictive models. The second block would be based on a series of deep learning models that implement different predictive theses and the third block is a delivery layer that includes the APIs and web interfaces used to consume the predictions.

This architecture model allows ITB to experiment with all sorts of different models in order to uncover predictive trends in crypto assets.

Why the Emphasis on Deep Learning?

Deep learning is one of those terms that quite often gets thrown around without too much substance so I would like to take a minute to clarify ITB’s emphasis in deep learning models. In general, deep learning refers to a type of machine intelligence architecture based on deep neural networks. From all the schools of artificial intelligence(AI), deep learning is one that has been inspired by the human brain. Deep neural networks are multi-layered structures that connect millions of neurons that work together to achieve an intelligent task( ex: predicting the price of Bitcoin). Although deep neural networks have been around since the 1970s, its only recent that they have become mainstream sparking an explosion in research and technology in the deep learning field.

In quantitative finance, there are emerging theories that believe that deep neural networks can be more effective at uncovering alpha opportunities than traditional machine learning or statistical models. In our experience, there are tangible benefits of using deep learning in crypto asset price predictions.

· Complex, Non-Linear Relationships: Deep neural networks can model very complex non-linear relationships between different predictors.

· Resilient: Through constant training, deep learning models have proven to be resilient to the constant fluctuations in the crypto market.

· Large Research Body: Deep learning research in quantitative finance is growing faster than any other discipline providing lots of research ideas that can be adapted to the crypto space.

At the same time, there are a couple of potential drawbacks of applying deep learning models to crypto asset price forecasts:

· Interpretability: Deep neural networks are complex and, therefore, really hard to interpret.

· Expensive to Build: Deep learning models are computationally expensive to build and maintain.

ITB Predictions on Ege Money

The current group of deep learning models are based on trades recorded in spot exchanges. The predictive neural networks have been trained in massively large datasets that include the trade history of some exchanges with clean order books.

Each prediction can have four possible results indicating price directions:

· Up: The price in the next hour is likely to trend upwards relative to the current target.

· Down: The price in the next hour is likely to trend downwards relative to the current target.

· Neutral: The price in the next hour is likely to stay within distance of the current target.

· Not Confident Enough: The model couldn’t arrive to a prediction with high confidence.

Should Predictions be Interpreted as a Trading Strategy or Recommendation?

ABSOLUTELY NOT! ITB price predictions are a directional indicator of a price movement. A trading decision and/or strategy involves many other considerations that should be evaluated by the party executing the trade.

Can predictions fail?

All the time! But ITB is completely transparent about successful and failed predictions so you would have complete visibility into the accuracy of the model.

How accurate are the models?

The current group of models have exhibited a very high levels of accuracy. Some of them have been tested for over 18 months. However, expect the accuracy to vary in different market conditions.

We are very excited about this partnership and hope that we can keep building value for crypto traders in Turkey through Ege Money. Please visit Ege Money and let us know about your experience with IntoTheBlock’s analytics and price predictions.