Mathematicians spend their time working to prove that a particular proposition is true. The work of data science is not only about the understanding of information in a particular business context, but it also involves the tradecraft of being able to manage data systems, search for new data sources, and find alternate ways of examining data to gain new and more effective insights. Perhaps there’s other data that can be used as a proxy for something that’s unavailable? For example, we can use social determinants of health to help inform our judgements about behavior and access to resources. That said, having a wider variety of data available and new ways of processing that data does mean data scientists have more options. You can’t fix bad data by adding more bad data. Thinking Outside the DatabaseĮven in the world of data science and big data, garbage is still garbage. If you can’t see the more effective way to find an answer given the business context, then you’re likely to either deliver uninformative results or fail to deliver any results at all in the time available to analyze and present results. Your organization’s data science capability has to be able to quickly understand new challenges, the operational business context and the data available in order to identify the best analytical approach.ĭata science requires you to not only able to apply a large number of modeling and algorithmic techniques from traditional statistics to machine learning, but also to know when to apply which approach. The work of building those new models and knowing which analytical techniques to apply, even in the age of machine learning, still requires specific knowledge and skills. In the mid-twentieth century, the complex computations required to build new business and scientific models were done by human beings that were referred to as “computers.” In the twenty-first century, the models have gotten far more complex and the calculations required to understand those models are done by electronic computers (of course). Balancing Data Science with Business Sense Without that healthcare knowledge, the rest of the data science analysis, modeling and solutioning process will have a critically weak foundation. A data scientist ignorant in healthcare culture and operations is likely to build an incomplete and potentially dangerously inaccurate mental model of the business situation being analyzed. A typical surgical team for a coronary artery bypass graft procedure, for example, will have more than 30 years of post-graduate education among them, not to mention the number of years of hands-on experience.The data science team working to optimize OR throughput doesn’t need all of the same education, but does need to understand what role each of those team members play and how their training impacts the way they work together to create the best outcome for the patient. The work of delivering healthcare requires years of specialized training. In order for your data science capabilities to be effective, the team needs to understand the terms being used, be familiar with how the data fits into the clinical and business operations and understand the higher-level business objectives that are related to those processes. Effective understanding and application of data requires contextual understanding.
More than most industries, the combination of unique scientific, business and technical terminology in healthcare is overwhelming. Healthcare industry and operations knowledge.Successful organizations balance several key components to create impactful data science capabilities:
The specialization and separation of duties that the data and analytics industry has built up over the past several decades has kept teams from collaborating together, using skills in an agile way and leveraging all of their available technical skills. More often than not, data science programs fail to create the necessary capability and fall short of the necessary impact. For businesses trying to gain as much insight as they can from data and create business agility, creating the data science capability is critical. Data science requires a special balance of industry knowledge, analytical prowess and technical skills that you don’t usually find in a single person. What makes the data scientist so rare and sought after is that we’ve spent the last several decades driving the specialization of skills in data analytics and creating distance where intimacy is necessary.
When the hype around data science started to grow, people began to claim that the role of data scientist would be the sexiest job of the 21 st century.
0 kommentar(er)
