Abstract:
|
We discuss statistical consulting in the age of cognitive computing, deep learning, and artificial intelligence. We begin by providing use cases from management consulting that demonstrate that the machine age is now, and is here to stay. We argue that industries (e.g., life sciences, banking, and insurance) now favor automated self-learning systems and analytics solutions over traditional statistical analyses. As a result, statisticians will need to adapt and embrace emerging technologies in order to stay relevant in an ever-changing consulting environment and data science community. Conversely, we posit that the blind application of artificial intelligence without core statistical principles, such as experimental design, sampling, inference, and causal analysis, comes with significant risks. We present use cases where traditional statistics were effectively used in consulting to augment the value of data science solutions delivered. We conclude by arguing that the hybrid approach of traditional statistics and its principles and modern machine learning is the future of data science consulting.
|