Background

Machine learning is a field of computer science that gives computers the ability to learn without being explicitly programmed. Arthur Samuel, an American pioneer in the field of computer gaming and artificial intelligence, coined the term “Machine Learning” in 1959 while at IBM. Machine learning is at the peak of its hype. Business leaders continue to evolve their Analytics and Business Intelligence programs. The post is about how to start a Machine Learning Initiative

Challenges

analytics2

Before we look at some things to do while starting a ML initiative, Let us look at 2 key challenges that undermine the same –

  • Data Analytics team continue to excessively focus on data collection methods rather than analysing and prioritizing business problems where ML can be applied which is where there is a real value of data.
  • Finding talent with good Machine Learning skills is the currently the top challenge. Data and analytics leaders often ignore tapping into existing talent by identifying people who have the right capabilities for ML.

No single tool or approach will be able to satisfy a fully developed ML initiative.

Developing a Machine Learning Model

data science

Business leaders are ready to start ML initiatives. Here are 3 main components for developing a machine learning model:

  • Training data — a cornerstone of ML. This data contains the information to learn from.
  • Learner algorithm — an algorithm (or a set of algorithms) to interpret the training data.
  • Output — a prediction or insight, derived from data.

To start with Machine Learning, It is important to start with answering some basic questions –

  • What are the right problems for ML?

    • The initial question which is worth solving is to understand what ML can enable. What are some specific problem which are definite and can be solved via ML.
  • Who will make ML successful?

    • Once you are ready with the right problems which the ML can solve, it is also necessary to have an ML development framework to structure the projects so you can identify tasks to achieve the results, and detect what you already have and where you need to fill in the gaps.
  • How do we start an ML initiative?

    • To start an ML initiative, you do not have to know how to build a model. The model could be your first ML model on which you are learning, or it could be a packaged application to solve a specific use case. The task here is to find someone who can take on the model development task.

Team learning and up-skilling should be an ongoing part of an ML initiative, not a finite task.

 

Facebookgoogle_pluslinkedinmail