1. As back-testing of trading signals take a lot of efforts, we expect the alpha version of CoinShell will contain only major trading pairs and most straight forward trading signals, such as moving average system. The benefit of moving slowly towards our goal is two-fold: it will make our workflow more efficient and it will assure that all the testings are done in the right way. However, we believe users will still have a lot of unique information at hand by using the alpha version.
2. One drawback of using machine learning models instead of pure human judgement when it comes to trading, is that when market environment changes, machine learning models need to change according to the change in environment. On the other hand, those models need to consume a relatively large quantity of data so as to stay reliable. Good news is, crypto market, despite being very young, trades 24/7 and generates much more data than traditional stock market in the same period of time. Banca will re-train the model parameters periodically.
1. We have finished the designing of user interface of CoinShell. Our frontend developer has started to work on the alpha version of the new application of Banca.
2. On the backend, we have realized the quality of data will decide 90% of the outcome of our models and we will allocate more resources to take care of that. After we are sure the data can meet the quality requirement, we will move on to testing for an optimized length of data that should be used for training the model parameters.
The BANCA platform serves the global cryptocurrency community. BANCA’s dynamic eco-chain uses AI and expert system that includes automatic management. The BANCA platform analyzes Big Data and delivers precise services tailored to the specific needs of our individual users.