Getting the most out of Machine Learning
“Since it is a long term investment, as long as an organisation has clearly stated upfront why they are embarking on their ML journey and what problems they’re looking to solve, it will be possible to compare the capabilities of a solution against what it needs to solve for.”
She notes that by using Machine Learning, enterprises are now able to better understand and process their data much faster using modern tools with established algorithms. “Potential outputs of a successful ML strategy can be powerful and measurable marketing campaigns or more efficient operations or logistics. Firms want to be able to utilise their valuable data, but not at a cost that is greater than its inherent value and Machine Learning allows for this, with the added benefits of more consistent decision-making and streamlined operations.”
Tucker says it is critical to ensure a problem is clearly articulated from the outset rather than attempting to retrofit a vendor or solution to an organisational challenge. “It is also important in the South African context to determine whether a ML system has been successfully applied to this environment. We have unique challenges and experiences that need to be catered for.”
The sheer volume and ongoing proliferation of data available means that organisations are now able to track the exact relationship between two seemingly unrelated things and predict a possible outcome. “Being able to answer these questions can benefit billions of people around the globe. If it is now possible to track causal factors that cause cancer, we can make more informed decisions around how to live our lives and can be a part of a long term solution simply by creating more data to feed into a ML solution to continue to improve the decision-making process,” she says.
In addition to being used as part and parcel of scenario-planning tools during the Covid-19 pandemic, Tucker says Machine Learning will continue to be deployed in almost every industry vertical as its many applications and benefits become more widely appreciated.
“As new verticals are created, so too are the opportunities for ML. 10 years ago ‘wearable devices’ were things in movies, and now these can predict heart attacks. Within healthcare, retail and financial services there are millions, if not billions, of data points created on a daily basis and ML can be used effectively to drive specific behaviour or provide a competitive advantage.”
