Machine Learning is computer science that gives computers the ability to learn without being explicitly programmed. Instead of writing lines of code with a lot of rules like “if something then do this”, Machine Learning software identifies patterns and creates algorithms based on large sets of data. It is an extremely powerful technique which is able to perform intelligent tasks which were up to now usually done by humans. It can recognize, predict, advice, optimize and classify, and therefore it can support the automation of various business and industrial processes, customer facing services and other data driven functions. Machine Learning is also an important part in the field of Artificial Intelligence.

Machine Learning is not new. It was already invented in the early sixties and in the past decades it had been used in the academic world and by large companies. However, it is one of the game-changing exponential technologies which only recently has become mature. With the growing availability of data, computing power and platforms, Machine Learning is now applicable within every organization.

It’s not a question of who has the best technology, but who has the best understanding and appreciation of what this technology can unlock.

Machine learning applications can be divided in four categories:

i. Cognitive machine learning

Cognitive machine learning is about recognizing and understanding text, speech, photo’s, audio and video. There are several examples surrounding us in our day-to-day lives like Apple’s Siri (speech recognition, text to speech), Facebook or Google photos (Face recognition), Skype (real time translation) or Shazam (audio recognition).

The same techniques however can now be used to support your business. For example: indexing documents, emails, images, social media, etc. to enable users mining and searching through them. Or process photos or audio files to measure and control the quality of industrial processes. Or use chatbots to communicate in natural language with customers in commercial or customer service, and even analyze emotions and social tendencies.

ii. Predictive machine learning

Predictive machine learning is about using (historical) data to gather a machine learning algorithm that predicts an accurate outcome for new cases. The technology is being used already by larger companies like Netflix (recommendation of movies you like, based on the viewing behavior of you and many others), Tesla (the self driving car software which controls the gas, brakes and steering wheels based on camera inputs and tons of data from other drivers and Philips (tumor detection based on radiology scans and historical data). Since the algorithms keeps learning from new cases, the software is getting better and better over time.

This technique is extremely powerful and can now be used to support your business with almost limitless possibilities. For example: diagnose mental health diseases and predict the best possible treatments, forecast revenues and sales for certain products in combination with weather predictions, predict the probability of customers moving to a competitor (churn rate) based on their contacts with customer service, predictive maintenance of engines and wind turbines based on data coming from IOT sensors, support intelligent credit management and collection processes by an automated advise to release or block a sales order, or supporting a legal office with advice based on jurisprudence data.

Applying predictive machine learning allows your organization to offer better services, be more efficient and outperform competitors.

iii. Optimizing machine learning

Optimizing machine learning is about applying algorithms to find the fastest, the best, the shortest, etc. outcome. In our day-to-day lives, we find these techniques among others in photo en video editing software applying color filters, but also in Route-planning software for finding the most optimized route.

It can be used to optimize various stages of your business or industrial processes like optimizing warehouse operations, optimizing the supply chain or cost comparison for the most effective healthcare intervention.

iv. Classifying machine learning

Classifying machine learning is an intelligent way to segment and cluster data in larger data sets. It is sometimes non supervised, which means its algorithm has not ‘learned’ from historical data but is smart enough to identify data on its own. Classifying machine learning is already being applied in various cases, for example in credit card fraud detection: If a credit card transaction takes place in Amsterdam and within 10 minutes the same credit card is used for a transaction in Singapore, the machine learning algorithm detects an anomaly and will block the transaction and warn the customer.
The same technique can be used in your business, e.g. to identify unusual behavior of servers in a data center to prevent service interruptions, or to classify customer or market segments, or to do crime analysis within the national police.