Machine Learning

Big Data

Machine learning is an artificial intelligence [AI] capability where computers have the ability to learn without being explicitly programmed. Using pattern recognition techniques, machine learning uses algorithms that enable the machine to learn from and make predictions on data. This is in stark contrast to strict analytics which relies on strictly static program instructions in making data-driven predictions or decisions.

Machine learning is one of the most significant changes in the the competitive landscape of business over the past decades. There is absolutely no question that AI will play an increasingly significant role in business competitiveness and sustainability over the coming years.

Although the fundamental capability around machine learning was developed in the early 1960’s, its value exploded with the combination of the Internet and massive computing power. This gave analysts access to big data sets, which provide data in sufficient volume needed to ensure that prediction probabilities are high enough to be relied upon.

The ability to use machine learning, and in particular the more recently developed deep learning techniques, to discover new relationships in these massive data sets, and to determine which relationships are the most important, amounts to an enormous shortcut in many kinds of work – product development, marketing, research, software development, algorithm performance optimization, even task allocation.

As the following examples highlight, in many large organisations, the benefits of the use of machine learning is already being proven.


Machine learning excels at the complex analytics that managers use to monitor activities under their responsibility, understand the root causes of issues as they arise, and accurately forecast future trends on the horizon.

Projected 2025 productivity gains = 45-55%.


Manufacturers currently use machine-learning technology to monitor, control, and diagnose faults in manufacturing plants.

Projected 2025 productivity gains = 30-40%

Professional services

AI in financial transactions already parse myriad news stories, financial announcements, and press releases, make decisions regarding their trading relevance, and then act in slivers of a second—faster and with greater information recall than any human trader.

  • Banks can use machine learning to detect fraud, finding charges or claims outside a person’s normal buying behaviour.
  • Futures Advisors use AI to offer personalized financial advice inexpensively and at scale.
  • Law firms are using computers to scan thousands of legal briefs and precedents to assist in pretrial research—work that would once have taken hundreds or thousands of hours of paralegal labor.

Symantec’s Clearwell system uses language analysis to identify general concepts in documents and present the results graphically. In one case, this software was able to analyze and sort more than 570,000 documents in two days.

Projected 2025 productivity gains = 45-55%.


How Machine Learning Works

Machine learning requires 3 large data sets: training data, test data and real data.

Using the training data set, an algorithm is developed to enable a machine to learn from the data – to identify patterns and predict future results with minimal human intervention.  There are both supervised and unsupervised approaches. The training data set must be a sufficiently large historical data with known outcomes, that can be interrogated by the algorithm to identify the probability of future outcomes based on similar data.

The output is then used again the test data to determine how good the analytical model is. Once this reaches an acceptable level of accuracy [high enough probability], the model can then be applied against real time data to predict the future probability of an outcome, such as a purchase or customer churn event.

This type of machine learning is currently widely used in marketing and fraud detection.

Expectations can often exceed the capability of machine learning. This largely because finding patterns is very difficult, with often not enough training data available. This is why China is emerging as a leader in machine learning and deep learning capabilities – having access to data from their higher population.