Smells Like Crypto Spirit
New Tax & Reports Made Easy + Data Science
Going Back to Our Roots
Ryan and Brian have worked together in the tax and accounting software industry together for nearly three years. Brian is going on with nearly six years of experience to bring this knowledge and consultation to every day traders. In crypto currency the world is murky for those that make thousands of transactions on exchanges and within other ecosystems. This experience helps leverage solutions for Vega into an industry that has many possibilities in technology and convenience. Implementing a free tool to users and our community conveys the message we have supported all along as a team which is deliver and deliver again sophisticated useful software.
Deploying What We Know Best — Taxes and Reports
We have done that with a stand-alone application that compiles and calculates all your transactions from multiple exchanges conveniently generating a single compiled ledger report in a spreadsheet and tax form rough draft!
In short and simply, a user takes their files containing transaction history from multiple exchanges and drags and drops them into the user interface. When ready, they send them through our service. The API services compiles and calculates all of those into a simple ledger showing gains and losses and even generates a rough draft of what your tax form will look like. Once those documents are completely generated, they are available for download for the user to store however they choose. Anonymously and serverless is the goal.
To see our demonstration live on stream as well as our other implementations of useful applications please see our YouTube live stream and videos.
Predictions: Drop a Little Data Science — From Ryan Schreck
Frequently Asked Question: Why use Recurrent Neural Network(RNN) for predicting cryptocurrencies?
Brownlee (2016) explains that time series is already a challenging task to perform but recurrent neural networks are here to the rescue! RNNs can handle very large architectures and are easily trained to deal with new situations when predicting values at time indexes. Recurrent neural networks are a type of artificial neural networks that allows it to consider previous data in the past. An example is price prediction for stocks or a cryptocurrency such as Bitcoin or Ethereum. A normal feed forward network is not as accurate and cannot consider past data. RNNs are adept at time series analysis and are also used for natural language processing for predicting the next word in a sentence. According to Ding et al (2015) we can use financial data from news sites and actual stock prices from the S&P. We can also apply their techniques to cryptocurrencies such as Bitcoin and Ethereum and correlate prices between markets.
The following diagrams show an example of an artificial neural network and a recurrent neural network.
Brandenburg (2017) explains that RNNs are recurrent because they can run the same computations for all the elements in the input sequence. We can use RNNs to analyze stock prices, forecasting and even predict words in a sentence when utilizing natural language processing. In figure 1 you see the x inputs and a loop back W on node S. As the RNN runs it is able to consider past time events so if we were at the fourth node St+1 we can consider past data from nodes St and St-1. We can train the neural network to perform the equivalent of a linear regression by utilizing a backpropagation algorithm through time such as gradient descent. The hidden layers use an affine transformation for the x inputs and weights and the RNN uses the LSTM (long short-term memory) to consider inputs and possibly forget them. Once the input goes through the activation function the output is sent into the next node and the process repeats.
Next Steps
The next steps for the Vega bot include implementing a full set of classes to handle one or more inputs to make predictions about Bitcoin and other cryptocurrencies. You have seen just a high-level overview of what is possible when we apply RNNs to time series analysis and forecasting. The fact that we can analyze social media data and correlate that with real cryptocurrency prices is a testament to the power of machine learning.
Agility on a New Level
Brian Carter has put in a full-time commitment and brings to the table something explicitly productive in the world of agile software development. Creating a new and exciting way to develop, test, deploy and bring quality of software up-front. A team of three people working under this Agile strategy is more effective than a team of ten using waterfall strategies. We plan to aim high and deliver on every level of agile development that is quality up-front.
In Conclusion
Vega is now working on parallel software applications and looking forward to implementing many more. The modular deployment of separate apps that are all able to utilize artificial intelligence just gives a sense of security and stability for promising technology. There are many plans set in-motion to work with AFIX to make this data high-performance and anonymous in tax reporting and sentiment predictions driving market decisions. Decentralization and IPFS is by far the best future approach to create any upcoming software ready for production-level.
Get Involved Today
Vega Intelligent Solutions is currently doing a crowdfund to enable them to raise some funds to complete their project. A minimum of $600,000 with less than a few days left to participate, visit the crowdfund page.
Our Vision and Mission
Bringing the Vega AI vision to reality requires support from the community and partners that believe in the concept. We are dedicated to win your confidence and stay vigilant as the project progresses. Vega A.I., no doubt, is a laudable project that must be supported. Funding could easily rake in a high opportunity in cryptocurrencies alike, because the value of Vega A.I. is not bound with limitation to create awareness and adopt the true A.I. potential. We are dynamic to implement integrated solutions for the future of software!
Resources
References
Brandenburg, J, 2017 “Applying Deep Learning to Time Series Forecasting with Tensor flow”
Figures retrieved from https://mapr.com/blog/deep-learning-tensorflow/
Brownlee, J. 2016 “Time Series Prediction with LSTM Recurrent Neural Networks in Python with
Keras
Retrieved from https://machinelearningmastery.com/time-series-prediction-lstm-recurrent-neural-networks-python-keras/
Ding,X. Duan, J. Liu, T. Zhang, Y. 2015 “Deep Learning for Event-Driven Stock Prediction
White paper retrieved from https://www.ijcai.org/Proceedings/15/Papers/329.pdf
