For this project, I chose the article entitled, “Why Businesses Fail at Machine Learning” by Cassie Kozyrkov. The link to the article can be found by clicking here.
The article serves to evaluate common misconceptions of machine learning through the utilization of a food preparation metaphor. She defines data as representing the necessary ingredients, algorithms as the “appliances” (or tools needed to accomplish the task), models as “recipes” (how to correctly use the algorithms to manipulate the data), and predictions as the final product, or “dishes.” See Figure 1 for a pictoral description.
Kozyrkov goes on to state that most “machine learning” courses only serve to teach students how to build appliances, and not how to correctly implement them. She advocates for an approach of Applied Machine Learning, in which data scientists become familiar with pre-existing algorithms instead of attempting to “reinvent the wheel.” Many professionals in the field start from scratch and build complex algorithms for tasks that have already been solved in simpler ways by others. Thus, the project can be completed more efficiently and in less time.
That being said, she concludes her article by reaffirming that the best analysis comes as a result of the collaboration of multiple individuals, each of whom are specialized in one area. She lists several occupations (Product Managers, Data Engineers/Analysts, Applied ML Engineers, Statisticians, etc.), and describes their ideal roles in a well-functioning project.
| Article.Likes.on.Medium | Article.Comments | Twitter.Shares |
|---|---|---|
| 17600 | 77 | 377 |
Having no prior experience in the fields of Data Science or Machine Learning, I thought that the article helpfully explained the nuances of these areas and how to best approach them. While it did not delve deep into specific techniques, the emphasis on a team-based approach with unique specializations reflected some of Dr. Wright’s comments in class. Likewise, the topic of “Applied Machine Learning” and the metaphor of food preparation made her argument easier to understand, even for a novice like myself.