Before you could only choose between two methods to produce time series predictions, classical algorithms and deep learning. These functions used a small amount of data to generate predictions making their forecasts inaccurate. We combine the two methods for more precise predictions.
Dyntell’s researchers are currently working to create an even more accurate prediction model built on new architecture. This adds correlation benefits to the ensemble process.
Through this process, we find that seemingly uncorrelated items can help make accurate predictions. You might not know why the dollar/yen exchange rate pattern appears in your sales data, or why your sales correlate with the country’s average daily temperature. But these similarities can help increase Dyntell Bi’s prediction capabilities–and give you greater insight into your future.