What is Data Science? To answer this question, we will compare two games with which nearly all of us are familiar: chess and poker. Both games are extremely complex, have numerous strategies for players to employ and defend against, and require careful planning and anticipation. The fundamental difference between these games is the fact that in poker, most of the information about the state of the game and how the other players are engaging is hidden. Chess, on the other hand, is played entirely in the open. In chess, the information is all there where anyone can see it, but it may not be clear to all involved.
At its core, data science is the process of opening information up and showing the previously unknown state. In the terms of our analogy, it is turning poker into chess. Done well, data science not only reveals hidden information, but also takes the extra step to equip players with knowledge they can apply to current and future games. The transformation from a concealed, unpredictable game to one that is open and informed is carried out via the scientific method. In practice, this process cannot discover all of the hidden states, becoming an exercise of risk mitigation.
So why is defining "data science" important to us? We believe the term is often used broadly to encompass concepts, including machine learning, database mining, and business intelligence. While data science can touch all of these, it is important to first describe the unmet need that data science addresses. Having a shared understanding of what data science is and is not will help every contributor start from a common foundation to solving complex health care challenges.
The “Science” in Data Science
Data science, with its scientific milieu, inherits the same set of principles that revolutionized the sprawling intellectual and philosophical movement of the 18th century Enlightenment: the scientific method. Scientists of the modern world, now bequeathed with a treasure trove of data and advanced computational tools, are able to harness this newfound capability to support teams and stakeholders in addressing a wide range of complex problems. Staying true to the method, they first establish clear research project goals and accordingly identify hypotheses and construct strategic plans of action. Then, they leverage their domain expertise to meticulously gather and prepare the necessary data. They undergo an iterative process of formulating, testing, and refining hypotheses, using experimental observations to excavate insights along the way. This rich knowledge base can then be harnessed to develop theoretical models that are able to reliably and efficiently resolve real-world problems when deployed to production. This is what we believe data science is, or should look like, behind the scenes.
Much of data science involves the construction of models that explain and predict phenomena we see in the world. A model is a mathematical way of formalizing our current understanding of a problem in a way that approximates reality. It typically describes patterns and trends in your data. The output of these models should guide intuition and facilitate decision-making or provide clinical or business intelligence.
Data Science in Practice
The connection between these three main topics shown in Figure 1 allows a data scientist to build stories based on what is revealed in the data, guiding them to decisions. It is, of course, quite difficult for any individual to be an expert in all three of these disciplines; it often takes interdisciplinary teamwork to be successful with complex data science challenges.
Figure 1. Data science is often represented as a Venn Diagram of three interlocking circles: computer science, math and statistics, and domain expertise. As shown in the intersections, software, data-driven research, machine learning/artificial intelligence (ML/AI), and cross-discipline communication skills are often necessary for the job at hand.
Data science teams have accelerated a variety of efforts around the world in research and industry. Some examples in health care include:
- Personalizing treatments and diagnosis with precision medicine, like predicting outcomes for treatment in breast cancer
- Automated diagnosis of serious medical conditions, like irregular heart rhythms
- Novel drug discovery, including new antibiotics
- Value-Based health care: Improving patient outcomes in community oncology practices
Next Blog Post Topic
The foundation of every data science effort is data. How data is acquired and identifying some of the challenges in the acquisition of data is the focus of our next blog post.
About David Hughes
David Hughes is the Principal Machine Learning Data Engineer for Octave Bioscience. He develops cloud-based architectures and solutions for surfacing clinical intelligence from complex medical data. He leverages his interest in graph based data and population analytics to support data science efforts. David is using his experience leading clinical pathways initiatives in oncology to facilitate stakeholder engagement in the development of pathways in neurodegenerative diseases. With Octave, he is building a data driven platform for improving patient experience, mitigating cost, and advancing health care delivery for patients and families.
About Octave Bioscience
The challenges for MS are significant, the issues are overwhelming, and the needs are mostly unmet. That is why Octave is creating a comprehensive, measurement driven Care Management Platform for MS. Our team is developing novel measurement tools that feed into structured analytical data models to improve patient management decisions, create better outcomes and lower costs. We are focused on neurodegenerative diseases starting with MS.