Predict Edge

Predict Edge is a web-based solution that allows you to relate price and fundamental data using AI and forecast prices with greater accuracy

Analyze the Data

With Predict Edge, you can analyze how every single variable in your data-set behaves and how it interacts with every other variable. These can be plotted on the fly, and can also be visualized graphically to get deeper insights into the correlations.

Build your own AI model

You can choose one of several AI models in-built into the solution and modify those as needed. The solution allows you to not just use any number of variables in your data-set, but also create your own interaction variables that the model can use to predict prices. The models include several advanced ML algorithms, some of which even take seasonality and trend into consideration while predicting prices.

Use in-built Back-testing Module

Once the model is built, you can use the in-built Back-testing Module to see the results. Further fine-tuning of the model can be done on the fly to improve back-testing results, and ensure that desired accuracy is achieved before applying the model to predict prices.

Solution Architecture

The Architecture for Predict Edge is as shown below. The Solution is flexible to allow you to pre-process the data, use the user-selected algorithm, fine-tune it and then generate decision-enabling results.

Benefits

  • Allows you to easily import price and fundamental data in several standard formats – csv, xml, json, Excel, etc. – thus ensuring a quick turn-around for setting up the solution.
  • You can visualize data using the latest technologies, including interactive and responsive 2D and 3D charts.
  • The Web-based, cloud hosted solution allows you to work from anywhere, using any device.
  • The choice of several pre-optimized algorithms allows you to pick the one that works best for you, and fine-tune it for best results based on your data.
  • Back-testing, Statistical Testing, and Analytics Modules allow you to get deeper insights into not just the interaction between the variables, but also into results, and how they can be used for trading / hedging.