Machine Learning can help businesses organize, interpret, and use big data by extracting meaningful insights and quickly solving complex problems. It can also help guide the region’s private sector towards fully harnessing the power that data has to offer.
Is data the new oil as some headlines have proclaimed? It certainly is a valuable asset, and in recent years, we have witnessed an exponential growth in the volume, velocity, and variety of data generated through technological innovation.
Unlike oil, the supply and applications of data are endless. According to the 2018 Annual Global CEO Survey conducted by PricewaterhouseCoopers (PwC), CEOs are struggling to translate vast amounts of data into better decision making.
This is where Machine Learning (ML) tools come into play. ML can help businesses organize, interpret, and use big data by extracting meaningful insights and quickly solving complex problems.
ML is a type of artificial intelligence (AI) where a computer learns from past experiences and can make predictions. For instance, humans label data, such as images of cars or cats, to teach the machine how to distinguish between them and later replicate the same classification process with future inputs of data (this is better known as supervised ML). Thousands of people from small-town China to big city India are making a living by teaching computers in this way. While this data labeling work may not seem like a job of the future, it is an essential building block for advancing AI systems.
You may also like:
Machines can also learn without any external or human guidance. For example, if a manager wants to segment their company’s consumer base from a large pool of individuals, the machine alone could find patterns in the data, which humans may not be able to detect (this is better known as unsupervised ML).
How are businesses in the region applying Machine Learning?
Most industries working with large amounts of data have recognized the value of ML technology. By gleaning insights from this data–often in real time–organizations can work more efficiently.
In the region, we have seen a large concentration of ML, and particularly AI startups, mainly in Argentina, Brazil, and Chile.
Some of these ML applications are probably more familiar to you than others. We’ve all seen those notifications that appear when we’re shopping online suggesting “other things you might like to buy”. In Argentina, online marketing platform Jampp uses ML tools to continuously analyze millions of app events and behavioral signals to help companies predict customer decisions.
Likewise, banks are using ML to prevent fraud. When you receive a text from your bank about a suspicious transaction, this is the result of ML algorithms detecting unexpected changes in your financial activity. Interestingly, a recent survey found that 83 percent of Brazilian financial service consumers would trust banking advice that was entirely generated by a computer.
Other ML applications may not be as well-known, such as making “fake” food. The Chilean startup, The Not Company (NotCo), produces healthy and sustainable food alternatives using ML algorithms that analyze the genetic properties of vegetables to create new products that emulate those produced by animals. These include a new recipe for mayonnaise, called “Not Mayo”, and alternatives to milk and ice cream.
Healthcare is also ripe for ML applications. In Brazil, LABDAPS, a laboratory at the University of Sao Paulo, is starting to conduct difficult diagnostics and mortality predictions using ML.
Finally, the Peruvian logistics startup Chazki is using ML to get around a common problem in the region: the absence of clear postal addresses. The company created a robot that learns the coordinates of delivery addresses and builds new postal maps, including destinations with no formal address. It has already expanded to the streets of Buenos Aires.
Machine Learning at IDB Invest
At IDB Invest, we’re also experimenting with ML, mainly to help us better capture and use the learning generated from our portfolio to design better projects.
For example, each project is evaluated at completion to determine if it achieved its expected development impact. Evaluation results have typically been manually classified into lessons learned by sector, country, and other categories. We’re using ML to automatically read the text documents and automate this classification process. We’re also working on using ML to make it easier for project design teams to find relevant information about past successes and failures in our data analytics system through automatic search tips.
The data revolution is here to stay. For the region to compete within it requires efforts on various fronts, from raising awareness about practical business applications of ML to building workforce technology skills and establishing laws and regulations to protect data privacy.
Just as machines need human guidance to learn, ML can help guide the region’s private sector towards fully harnessing the power that data has to offer.■
SUBSCRIBE AND RECEIVE RELATED CONTENT