Hot money is frequently risky money. ICO is part of it. One may read the white paper, but it is always time-consuming with nothing but beliefs and rules of thumb hoping to find the right ones. Over the past years, hundreds of coin offerings were launched with little information past the ICO period.

With AI, it is possible to evaluate the probability of success 

The first approach, easy to apply, if the logistic regression. Behind this fancy term hides a powerful mathematical tool. Identify a list of key parameters (e.g., number of people on the whitelist, topics covered by the project, expected use cases, average amount invested, etc.). Then apply the logistic formula on it to obtain a “boolean” score. More simply: will the investment be profitable or not. This approach is factored sensitive. A long and hard job is required to realize this “old fashion” statistics approach.

Machine learning for further

The idea is to start from data; not from a hypothesis. Collect as much data as you can on every ICO available on the market. To be statistically sufficient: a minimum of 30 ICOs is required. For each, collect all available variables. Do not raise a hypothesis. Run a neural network on it, e.g., a convolutional one. The model trains itself until he finds a solution to the problem: how can I make a high return on ICO investment?

Now, take a position

Now, and it is the hardest part: validate your model with existing data. If it works, apply to the next ICO you want. The predictors, chosen by the model itself, will tell you whether or not to invest.
All that will be possible with NeuroChain. Until then, it may be difficult to develop a model, link it to the data, train it and get a result.
Invest smart. Invest NeuroChain.
Information about the ICO date is about to be released. Stay tuned, join us on Telegram.
Photo credit: Splitshire

Leave a Reply

Your email address will not be published.

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>