- Machine Learning
COVID-19 was a major pandemic that struck the world at the beginning of the year 2020. Many companies suffered sudden disruptions in their manufacturing operations, logistics and even in their capacity to reach their customers. This capstone project addressed the need of a global pharmaceutical company to understand what digital capabilities were required to be more resilient. The research team conducted in depth interviews and reviewed the literature on resilience: it was identified that transparency and advanced analytics are the main digital capabilities that can increase resilience. Then the research team implemented machine learning techniques to demonstrate how the utilization of advanced analytics can improve resilience. For this analysis, the research team implemented decision trees and random forest in two different datasets from 2019 and 2020 to draw conclusions about what influenced the company’s ability to fulfill orders under a normal state and a disruptive state, as a measure of resilience. The results of this analysis showed that order quantity, location population, and the category of products by level of sales are features that help determine the potential disruption of fulfillment orders; this knowledge can help increase resiliency.