Diversified Predictive Analytics 

There is no singular-best approach to predictive analytics - which is why so many “experts” focusing on one approach will get “out-forecasted” by other approaches. The optimized approach is to combine all 4 meta-approaches: Qualitative, SME, Basic Modeling, and Advanced Analytics/Machine Learning. Essentially, leveraging a series of approaches, methodologies, and models (i.e. ensemble modeling) assures all possible inputs, influencing factors, assumptions, drivers, possibilities, parameters, technologies, weights, data sources, algorithms are enjoined and synergized into a Holistic Optimized Approach - which will produce the lowest absolute margin of error of any single approach.