Quick Intro to Machine learning for beginners.

When researching Machine learning, things such as natural language processing, conversational AI, computer vision and reinforcement learning are often talked about. But its extremely hard for beginnings to even grasp what Machine learning is and how we can take advantage of it within our applications. I aim to help beginners understand ML and understand its potential to improve a company’s productivity by up to 60%.

Javascript is the programming language of the Web. It is by far the most advanced and used programming language. It has become extremely advanced over the years. We can codify machine learning principles with the help of a few Machine learning libraries such as Brain.js and ConvNetjs. This is advantageous for Web Developers as they won’t have to use Python the language most people tend to think about when talking about Machine learning.

Machine learning is one of the applications of Artificial intelligence that enables a system, or web service to automatically learn and improve the user experience without having to code such a behaviour manually. Machine learning uses data to learn and process different models. For example, it takes a range of different football wins and losses to determine if a specific team says Manchester United has a chance of winning based on past metrics with the opposing team.

Machine learning aims to give computers brains and enable them to learn the best behaviours and actions based on data collected about a specific task. Typical computing involves a programmer having to code a specific behaviour into a system manually and performing the same computation over time, whereas with the machine learning model, the system collects data and begins to learn and improve the system over time. For example, the more shows you watch on Netflix the better its suggestions algorithm becomes over time.

2 Types of Machine learning methods

There are 2 types of Machine learning methods, Supervised and Unsupervised learning methods. With supervised, the developer must pass labelled data to the algorithm often referred to as training a data set. For example, Supervised learning maps input and expected output. We must pass a complete data set to the ML model. Once passed the Algorithm will use that data set to make further predictions. An example may be predicting if a person is smiling from a data set of inputting a smiling man and tagging it smiling. The algorithm will determine most likely that a smiling picture to the tag smiling.

Unsupervised machine learning is when there is no classification of data and the data is not labeled when being trained. The system would have to infer ways to describe structures of unlabelled data. This type of Machine learning is often done with minimum human supervision. The two main uses of this type of learning is for Cluster analysis and principal components. Cluster analysis is the act of grouping a set of objects in the same group. For example, a range of animals as Mammals based on attributes available in the data.

Common uses of Machine learning

A common use of Machine learning is Facebook’s news feeds. It analyses its users’ behaviors such as if you scrolled past a certain post or liked a specific post. It uses these data to train a Machine learning model and help it make predictions on the types of post its users are interested in and displays them to the user. Data attributes such as Like, dislike and comments are used to make such predictions.

HR can train Machine learning models with the best attributes to look for in an applicant and make predictions automatically if a person applying for an open job is likely to be a perfect fit. This saves time for both parties as the HR team doesn’t have to move forward with interviews that are likely to not lead to a Job.

I hope this helps you understand the very basics on Machine Learning. In a separate post I will be going over Brain.JS and how to easily use to create Machine Learning applications on the Web.

Originally published at https://www.linkedin.com.