Decoding Machine Learning
Machine learning is, probably, the buzzword in almost all the industry. Be it IT, pharmaceutical, retail, you name it. People have started boarding the ship of machine learning without prior knowledge. I feel this is good and bad. But, with this piece of knowledge, I will throw some light on machine learning, especially for newbies.
Welcome to the world of machine learning. It’s easy to see why the word is making much noise around it. Let me express my views on machine learning in the simplest of the literature. Machine learning is a field of computer science that gives computer systems the ability to “learn” (i.e., progressively improve performance on a specific task) with data without being explicitly programmed (Wikipedia).
I divide machine learning into two broad categories 1) Supervised ML (Machine Learning) 2) and Unsupervised ML. There are other pre-processes involved in data mining, data manipulation, etc. In the categorization, I have classified the algorithms.
Supervised Learning: In this type of algorithm, we supervise (learn) the previous records of the data to understand the demographic, behavioural and/or transactional features of the individuals/users. This data is then used to predict the values in the future, assuming all the conditions are the same.
Don’t we do this in real life? If we friends meet up and have that one friend who always says no to a movie, we might not ask him again. Using his previous behaviour, we predicted his answer and didn’t bother to ask him. Organizations across the globe use these algorithms to identify potential customers and retain existing ones.
Unsupervised Learning: In this type of algorithm, we do not predict the values, but basically, try to identify similar patterns of different individuals and club them into one. Clustering is the most straightforward example of this.
We know that two of our friends in the group like soccer (Football), three like cricket and the other two like snooker. We have categorized our friends into these groups using their love for the sport. Using these models, companies float their offers to prospective customers, an essential part of marketing analytics.
The majority of the algorithms can be classified into these broad categories. I would suggest all the newbies learn and understand the involved science. This will make the game even more enjoyable.