Unlike other hedge funds, Numeraire does not want to rely on an individual or a team of financial market experts. Normally, they would invest a large amount of capital, in order to allow the capital to grow. However, this could also be solved with machine learning. A computer can learn on the basis of large data sets how the money could also be invested. That’s exactly what Numeraire focuses on.
To implement this concept and improve the investment strategies using it, Numeraire wants to attract more data scientists. Numeraire aims to generate the best possible returns using better models that can make predictions for the stock market. Of course, this should also benefit investors.
- 1 What is Numeraire?
- 2 The problem that Numerary solves
- 3 The Numeraire team
- 4 How does Numerary work?
- 5 Numeraire’s security
- 6 The Numerary Token: NMR
- 7 What is Numeraire all used for?
- 8 ICO
- 9 What is Machine Learning?
- 10 What is the difference between Machine learning and Artificial intelligence?
- 11 Numerary (NMR) Chart
- 12 How and where can you buy NMR?
- 13 Conclusion
What is Numeraire?
Numeraire is a San Francisco-based blockchain startup founded by 29-year-old South African Richard Craib in 2016. The startup is financially backed by Union Square Ventures, AngelList founder Naval Ravikant and First Round Capital.
Machine learning algorithms (also called artificial intelligence) have solved several major problems in recent years. Think, for example, of recognizing handwritten numbers in the MNIST dataset. When the dataset became available in 1998, AI algorithms were still very imprecise. At the moment, many companies are working on further developing AI. After all, it is a way to automate more processes and to take over the work of people. And that is exactly what Numeraire is doing.
As we mentioned in the introduction, Numeraire is a platform that allows developers and data scientists to experiment with machine learning. They can do this with the different datasets and models that Numeraire offers on the platform. The goal of the platform is to bring more decentralization to the data science world. Developers can then access datasets much more easily. See a short video below explaining what Numeraire actually means
The problem that Numerary solves
It is sometimes very difficult to get data. Suppose a developer wants to develop an application for which he will use machine learning. Such an application could help a lot of people. However, a developer does need enough data to be able to apply AI. A developer could collect that data itself, but could also get it from somewhere.
When the developer chooses to collect data himself, it takes a very long time before enough data is available. But when the developer chooses the second option, he has to trust that the data is also reliable.
By decentralizing a dataset, it is much easier for developers to develop machine learning. And that is exactly the goal that Numeraire supports. They would like to see machine learning and AI further developed because they believe in the idea of machine learning.
The Numeraire team
Numeraire’s currency, NMR, refers to a product of the Numerai company. The board of this project consists of Richard Craib, Georey Bradway, and Xander Dunn with Joey Krug. The company was founded in February 2017 and is still very new to the market. Numerai announced in September 2017 that it had exited Bitcoin and was phasing out Numerai. All payments were converted into Ether and Numeraire from then on.
How does Numerary work?
In the process of Numeraire, the data experts participate in a kind of competition for the best forecasting model. They calculate forecasting models using coded data series that represent stock market information.
To that end, the data scientists create an algorithm to discover the right data patterns. Then the capital market and stock price forecasts are uploaded to the website. So far, Numeraire has received 40 billion such forecasts, and currently Numeraire receives 100 million forecasts per day.
The relevant prediction models are ranked, with the top 100 data experts being paid in Bitcoins, Ether or Numeraire. The idea behind this is that the better the predictions or the higher the ranking of the data expert, the more Numeraire they will receive. So this encourages the developers to develop the best possible model. This is how they can earn the most money, and that also benefits the platform.
However, these competitions are not entirely risk-free for the data scientists. Participating data scientists are required to deploy a portion of their numerical value for a specified period of time to test the effectiveness of the prediction model. So it uses strike.
If the data scientist is wrong with his forecast, the NMR coins will be destroyed, which should ensure the quality of the forecasts. If the data scientist is right, he will get Numeraire back and Bitcoins as an extra reward. Currently, the payout is capped at $15,000 per month, meaning top experts can currently earn up to $54,000 per year. A total of $200,000 has already been paid out to data scientists.
Data is incredibly valuable. There are also a lot of hackers who see this as prey. That is why it is important that Numeraire’s data is properly secured. And they do. Since Numeraire uses the ERC-20 token, the network is secured using the Proof of Stake (PoS) algorithm. The block validators are thus selected based on the size of the stake they bet. In this case, those are the data scientists, who will have to bet a stake if they want to participate in the competition.
Proof of Stake is an algorithm that you often see in blockchains. After all, it is the second most popular consensus algorithm after Proof of Work . The advantage of Proof of Stake compared to Proof of Work is that PoS is much more sustainable. PoW uses a lot of computing power, which also costs a lot of energy. Numeraire is therefore a fairly sustainable project. Users who want to participate bet a stake, and based on the stake size, they are selected to add blocks to the blockchain. So there are no users engaged in providing work that is ultimately not used, as is the case with Bitcoin.
The Numerary Token: NMR
The company’s core idea is a cryptographic token that can be used in a new auction mechanism. In this way it wants to make the oversupply economically irrational.
Numerary is an ERC20 Ethereum token. These tokens are called smart contracts that run on the Ethereum blockchain. All posted Numerary transactions are sent directly to Numerai. The Ethereum smart contract dictates that there will never be more than 21 million Numerary tokens in the market. After the initial distribution, the smart contract will form a fixed number of Numerary each week until the maximum is reached. It is quite remarkable that there are only 21 million Numeraire available. There are not many projects left with such a low total supply. According to many, this only benefits the value of NMR. The lower the supply is, the higher the price of the token will be. After all, this only applies if the demand for the token does not fall. If it does, the value of NMR will not increase immediately.
What is Numeraire all used for?
Numerai wants to use this coin to build an artificial intelligence hedge fund. Numerai’s auction mechanism consists of competitions, each with a prize pool that is matched by a certain number of coins. The prize is awarded to data scientists through an auction mechanism. This auction mechanism leads to balanced bidding behaviour, demonstrating that rational data scientists have confidence in the special capability of their own models.
These capabilities are based on getting good results on new data. This particular mechanism also provides pristine and natural arguments for the economic value of a numerary token. By successfully completing Numerai’s machine learning competition, data scientists can continuously win Numeraire.
A so-called Initial Coin Offering ( ICO ) has never taken place. Small investors could therefore not participate in the project in advance. To encourage data scientists to submit their data models, Numeraire launched its own cryptocurrency: Numeraire (NMR). This is a cryptocurrency based on the Ethereum blockchain. The maximum number of coins (total supply) of Numeraire is 21 million. That is just like with Bitcoin.
Initially, 1.2 million Numerary tokens were issued to a network of 19,000 data scientists. New Numerary tokens will be issued weekly to data scientists. These will be used to help improve stock market forecasting models in the future.
What is Machine Learning?
While Machine Learning is not new, its precise definition is still confusing to many people. In concrete terms, this is a modern technology that makes it possible to discover patterns and make predictions from data. This is done on the basis of statistics, data mining, pattern recognition and predictive analytics. The first algorithms emerged in the late 1950s. The best known of these is the Perceptron.
Machine Learning is very effective in situations where insights need to be discovered from large sets of disparate and changing data. This data is also known as Big Data. For the analysis of such data, it proves to be much more efficient than traditional methods in terms of accuracy and speed. For example, based on information about a transaction, such as amount and location, and historical and social data, Machine Learning can detect potential fraud in a millisecond. This method is therefore significantly more effective than traditional methods for analyzing large amounts of data.
Machine Learning can be defined as a branch of artificial intelligence that includes many methods of automatically building models from data. These methods are in fact algorithms.
A traditional computer program performs a task by following precise instructions, and thus systematically in the same way. A Machine Learning system, on the other hand, does not follow instructions, but learns from experience. As a result, performance improves as the algorithm is exposed to more data.
What is the difference between Machine learning and Artificial intelligence?
The difference between machine learning and artificial intelligence is very small. Machine learning is part of artificial intelligence. Machine learning is all about making a machine learn from the data it receives. However, there are also other forms of artificial intelligence. The goal is always for a computer to get smarter and solve difficult problems, the same way humans do.
Numerary (NMR) Chart
How and where can you buy NMR?
You can buy Numerary (NMR) on various crypto exchanges. Think of the large exchanges such as Binance and ByBit, but also of the smaller and decentralized exchanges. It is also possible to buy Numeraire on, for example , Sushiswap , PancakeSwap or Uniswap . You can buy NMR with any cryptocurrency.
Since Numeraire is an ERC-20 token, you can store it in almost any wallet. This is the most commonly used token. If you have a lot of Numeraire in your possession, it is recommended to store it in a cold wallet. This is also called a hardware wallet , and is much more secure than a hot wallet, which is always connected to the internet.
Numerai wants to use its cryptocurrency Numeraire to pay data scientists who will ultimately yield better returns. Whether big data and predictive models based on artificial intelligence will actually drive better returns in the stock market is not yet entirely certain.
Numerai has yet to prove that the concept works. In theory, using the power of data, the hedge fund should outperform, causing more capital to flow to the hedge fund.
This in turn should lead to more revenue for Numerai’s partners and thus higher payout percentages to the data scientists, which should also increase the value of the cryptocurrency Numeraire. Investors who have invested in the Numerary token can also benefit from this appreciation in value.
However, investors should keep in mind that cryptocurrencies fluctuate widely in price and the market is heavily influenced by speculators. Rapid price gains, but also high price losses are the order of the day. Therefore, investments in cryptocurrencies are generally only suitable for speculative investors.
Thanh Lanh Tran(1989) is Chief Editor from BitcoinUSD.com