![]() ![]() ![]() The process is actually quite similar to that of a human analyst or investor. Whilst AI is not a crystal ball (yet), it is very capable of making forecasts and decisions using powerful quantitative algorithms given the right data inputs. But is it really possible for AI to predict the Stock Market and get rich? In more recent times, with the major breakthrough in AI in the form of Deep Learning (machine learning inspired by the structure of the human brain), open source frameworks such as Tensorflow, and open communities such as Kaggle, it has become possible for anyone to easily learn, build and share powerful predictive models.Ĭombining these breakthroughs with cloud computing, which allows us to train our predictive models with unprecedented amounts of data, and the predictive power at our fingertips is greater than ever. Successful investment algorithms, however, were still quite complex and very much a 'dark art'. In the 70s Box and Jenkins created the ARIMA mainframe algorithm while in the 80s the PC revolution saw Excel spreadsheets becoming popular among traders and quantitative funds. Ancient civilisations resorted to various forms of divination and primitive record keeping such as Oracle Bones and Quipu knots to cater for forecasting needs ranging from weather and farming to commerce and military conquest.įast forward a few millennia and data collection and record keeping has increased in sophistication from bones, bamboo, stones, paper to digital storage and display. The desire to foresee future events before they unfold has gripped our ancestors since the dawn of time. Part 1: Forecasting and AI A brief history of forecasting techniques I would like to give a massive shout out to Quantino team members: Lucas Rafagnin, Ira Cohen (co-founder Anodot), Sami Raines and Raymond Au, without whom none of this would've been possible. More importantly, it might demonstrate the possibility of using powerful high level AI services to solve business problems without necessarily having teams of data scientists on hand. Hopefully, this article will also inspire you and your organisations to look into the many practical use cases for AI that are already emerging among leading enterprises. Our app is hosted at quantino.ai, slides below: We managed to achieve predictions up until November 21st with < 1% Symmetric Mean Absolute Percent Error (SMAPE). It was a lot of fun and we were quite surprised at how easy it was to create a responsive AI application in such a short period using AWS Serverless and SaaS APIs such as Anodot. ![]() Team Quantino demonstrated a stock forecasting application for predicting the stock price movements of all four major Australian banks over a period of two weeks, which we built in two weeks. On 7th of November, Team Quantino (mashup between Quantitative and Contino) participated in the Melbourne Data Showdown 2018 event, hosted NAB at their fantastic venue over in The Village Docklands. “Prediction is very difficult, especially if it’s about the future.” Yun Zhi Lin 12 December 2018 Building a Stock Forecast Application in 2 Weeks Using Serverless AI and Responsive Design ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |