Top High Impact Machine Learning Applications by Industry
By David Weedmark2021-11-047 min read
In the last decade, machine learning (ML) models have made organizations all over the world and in every industry more productive, more profitable and better able to serve their clients.
Machine Learning Applications by Industry
Every organization has its own challenges when it comes to developing and adopting machine learning into its operations. But most of those challenges have been seen before either in their business sector or outside of it. By borrowing best practices from leaders in data science, organizations can rise above the competition and become model driven.
With the assistance of machine learning, financial institutions can offer more personalized services, products and recommendations to clients. For example, a global Fortune 500 firm uses ML to improve customer experience with call centers and improve customer communications. They also use ML to prioritize recruiter workloads in HR to improve their success with hard to fill job openings.
ML is also being used by financial organizations to identify anomalies that can represent risk, like unusual trading patterns, which can alert risk managers to problems or trigger an immediate resolution. Moody’s Analytics’ suite of RiskCalc models is used to estimate the probability that a firm may default on its debt. Used across the globe, these models are based on industry standards but customized for different situations, taking into account regional variances as well as different risk tolerances required for different insurers and lending institutions.
When properly developed and deployed, machine learning models can be a perfect match for an insurance company’s underwriting needs. Model-driven insurance firms are disrupting the insurance market and increasing market share at the expense of slow-moving incumbents.
Being a model-driven insurance company means the overall risk profile will be significantly affected. However, as the insurance branch of BNP Paribas Cardif has demonstrated, this impact can be very beneficial if models are developed, deployed and monitored under an umbrella of strong governance. With over 10,000 employees in 35 countries, serving 100 million policyholders worldwide, BNP Paribas has deployed ML/AI projects across the organization, not only increasing the productivity of model development but building credibility even among its most risk-aware stakeholders.
Another insurance provider, TopDanmark, was able to use Domino’s MLOps platform to develop and deploy ML-based apps to significantly improve the speed and accuracy of its claims processing, allowing 65% of cases to be automated. Today customer approval is granted in 1-2 seconds, nearly 800 times faster than with a human adjuster.
Media and Technology
The media and technology sector has many opportunities for data science impact, Yet the industry has its own challenges, most of them caused by its greatest strengths — the internet, dispersed workforces and constantly changing technologies, to name a few. Keeping teams onboard is particularly difficult, with a retention rate of only two years. Having a state-of-the-art platform with access to cutting-edge technologies and built-in collaboration tools has been an indispensable asset for hundreds of tech and media clients.
In one famous case study, Red Hat used Domino’s products to unite its globally diverse workforce to create machine learning models that improved revenue forecasting with more efficient sales analytics while improving the efficiency of the company’s data science community 10-fold.
Health and Life Sciences
Diverse global workforces and the need to protect sensitive intellectual property are just two of the many challenges faced by data science teams working in the health and life sciences sector. The enterprise-grade security precautions and rich collaborative features built into Domino have helped some of the top companies produce winning ML-based apps to build on innovative solutions and research.
Bayer recently launched a multiyear research pipeline to develop seeds that improve crop productivity and protect them from insects and herbicides. Evidation is using Domino to accelerate model velocity, with production grade models implemented in as little as 8 weeks, helping clients manage their health better so they don’t become patients.
Janssen, a division of Johnson & Johnson, is training deep learning models 10 times faster with Domino and Nvidia GPUs to accurately diagnose and characterize cancer cells through whole-slide image analysis. This is allowing them to increase the number of patients screened for clinical trial eligibility.
For years, the manufacturing sector has lagged behind other industries with ML adoption due to complex operations, numerous regulations and unwieldy logistics. These are still challenges today; however, numerous manufacturers have succeeded with predictive and prescriptive models that streamline inventory management, optimize marketing spend and help determine optimum pricing and promotions.
Leaders in the market are looking beyond operational applications to ground breaking machine learning use cases. Just last year, Lockheed Martin successfully built an ML-based app that simulated dogfights against an F-16, beating the human pilot in all five engagements.
Impact That Machine Learning Has on Society
Machine learning has proven itself to be a tremendous benefit, not just to business and consumers but also to society as a whole. ML models help diagnose illness and cancer, predict traffic patterns and keep children and neighborhoods safe, among a multitude of other benefits.
In one noteworthy example, Audubon has used Domino’s data science platform to create 180,000 models that analyze over 140 million observations to predict the impact of different climate-change possibilities on bird species.
The Climate Corporation employs about 200 data scientists and data engineers using Domino’s data science platform. They are using machine learning, deep learning and other sophisticated methods of analysis to help farmers grow better crops on an individual basis. These models are able to determine what seeds or hybrids to plant, how to plant them, how to cultivate them and when to harvest the crops to maximize their yields.
Machine Learning in Action
If your organization has yet to achieve the success with its machine learning applications that these companies have, perhaps it’s time to revisit the tools you are using. Twenty of the Fortune 100 companies now use Domino’s enterprise MLOps platform to drive their machine learning projects. Whether you work in R, Python or MATLAB, Domino has the libraries and collaboration tools you need.
David Weedmark is a published author who has worked as a project manager, software developer and as a network security consultant.
Subscribe to the Domino Newsletter
Receive data science tips and tutorials from leading Data Science leaders, right to your inbox.