Abstract: Systems and methods are disclosed for predicting the performance of a financial instrument by extracting influential features and emotional sentiment data from received data values, common features, and internet sources, each corresponding to a financial instrument, building multiple predictive models using random selections of the data values and combining the predictive models to create a combined prediction model. The combined prediction model provides a combined predicted data value where the combined predicted data value is a performance indicator, such as a classification count or percentage, of data values that correspond to the financial instrument and classify correctly via the prediction models.
Abstract: Systems and methods are disclosed for predicting the performance of a financial instrument by extracting influential features and emotional sentiment data from received data values, common features, and internet sources, each corresponding to a financial instrument, building multiple predictive models using random selections of the data values and combining the predictive models to create a combined prediction model. The combined prediction model provides a combined predicted data value where the combined predicted data value is a performance indicator, such as a classification count or percentage, of data values that correspond to the financial instrument and classify correctly via the prediction models.