Introduction to Machine Learning

We, humans, can easily differentiate between objects and recognize different patterns. It is an effortless process. Why? because we have been learning these things since our childhood. Human learning happens in three ways: Learning by self, Learning under expert guidance and Learning guided by previously gained knowledge.

But Machine learning is different from Human learning. There are multiple ways of defining Machine Learning, but the most relevant and precise definition accepted universally was stated by Tom M. Mitchell, Carnegie Mellon University

” A Computer Program is said to learn from experience E with respect to some class of tasks T and performance measure P, if its performance at tasks is T, as measured by P, improves with experience E”

In simple terms, it essentially means that a Computer is said to be learned if it improves it’s performance based on past experience.

Machine learning is an integral part of developing a better future. In due course of time, Machine learning along with AI will change the course of development. It holds the potential to transform the way we carry out our daily tasks.

Everything in machine learning revolves around data. As technology continues to evolve, we gain more and more useful and relevant information which is used by industries. Industries adopt machine learning to improve their processes and gain insights from data.

Here are a few industries that make use of machine learning to boost their performance:

• Healthcare : Machine Learning models helps doctors make better decisions and helps develop a better personalized plan. It helps to predict patient outcomes, diagnose diseases, analyze reports and detect early warnings. This in-turn reduces healthcare costs.
• Finance: Predict market trends, analyze risks and fraudulent transactions. Improving business efficiency by automating trading decisions and optimizing investment decisions for trade.
• Marketing: Understand customer behavior, analysis of market trends ways to improve customer interaction. Help devise new advertisement strategies and pricing policies
• Production: Optimize manufacturing chain, predict risks and help in quality control. Automating routine tasks.
• Education: Enhancing personalized learning methods. Identify areas of weakness and provide improvement strategies. Help mentors plan efficiently in order to meet every students needs.

Now that we know why Machine learning is important let us understand how machines actually learn. For better understanding we can divide the process into three parts:

• Data Input: The collected data or information is utilized by the machine learning model. But it is important to note that not always do we get our data in the required format. We need to extract only the necessary information and store it in a suitable format.
• Abstraction: The process of summarized knowledge representation of the raw data. It allows the algorithm or model to work more precisely with complex data sets.
• Generalization: This process helps to tune up the abstracted knowledge. If we try to overfit a model for some data set, we cannot predict values correctly for the remaining data sets. Hence generalization is important.