Chapter 14: Combining KNIME and H2O to Predict Stock Prices

In every book about time series analysis, there must be at least one example application to predict stock prices, that is, the forecasting problem. So, we conclude this book with a final chapter describing a forecasting application and the integration of KNIME Analytics Platform with H2O, which is another open source platform.

The stock price prediction problem is infamously difficult to reach accurate results for as the data changes quickly, on a daily basis. Furthermore, the drivers of these changes vary from physical factors, such as environmental disasters, to socio-economic factors such as political elections, and even to random factors that cannot be predicted. Thus, we're dealing with data with complex structures and interrelationships, which, as a result of the increased number of exchanges in the globalized stock market, is produced at a high frequency and processed in real time.

At the same time, the...