What is Machine Learning?

Here is everything you need to know about Machine Learning.

The world is filled with data. Lots and lots of data. Everything from words, spreadsheets, videos, pictures, music, and more. It doesn’t look like it is going to to slow down anytime soon. Machine learning brings the promise of deriving meaning from all of that data.

Data lies around us

For centuries, human have analyzed data and adapted systems to the changes in data patterns. However, as the amount of data increases and surpasses human's ability to manually control and write rules for it, we also shift to use automated systems that can learn from the data and changes from it to adapt to a shifting data landscape.

Machine Learning is already everywhere

Nowadays, we see machine learning in all applications around us, but it isn't always apparent to us that machine learning is actually behind it all. Let take an example. While identifying between cats and dogs pictures, or tagging people and objects in images are applications of machine learning, some features like video recommendation systems are also powered by machine learning that you may not even realize. Search engine like Google or Bing is also a big example of machine learning. Every time we search for a keyword, we use a system that use machine learning at its core, from understanding the text of your query to configuring the result based on your personal data collected.

• For example, if you search for "cookies," machine learning determine which result to show first, depending on whether it learns you are a web developer or a cookie lover. Maybe you're both :)

Bing Search

Today, machine learning’s immediate applications are already quite wide-ranging, including image recognition, fraud detection, recommendation engines, as well as text and speech systems. These powerful capabilities can be applied to a wide range of fields, from diabetic retinopathy and skin cancer detection to retail, and of course transportation, in the form of self-parking and self-driving vehicles.

Waymo, a self-driving car startup

An expected feature

It wasn’t that long ago that when a company or product had machine learning in its applications, it was considered phenomenal. Now, every company is looking to use machine learning in their products. It quickly becomes an expected feature. Like when we expect companies to have a website that works on our mobile device five years ago or a dedicated app for their products, the day will soon come when it will be expected that our technology will be personalized, insightful, and adapting.

As we use ML to make existing human tasks better, faster, or easier than before, we can also look further into the future, when ML can help us do tasks that we never could have achieved on our own.

An understandable and concrete definition

As we learned about ML, we need some kinds of definitions that encapsulate the ideal objective or the ultimate goal of ML. But first, let's take a look at some practical definitions from reputable sources:

1. “Machine Learning at its most basic is the practice of using algorithms to parse data, learn from it, and then make a determination or prediction about something in the world.” – Nvidia 

2. “Machine learning is the science of getting computers to act without being explicitly programmed.” – Stanford

3. “Machine learning is based on algorithms that can learn from data without relying on rules-based programming.”- McKinsey & Co.

4. “Machine learning algorithms can figure out how to perform important tasks by generalizing from examples.” – University of Washington

5. “The field of Machine Learning seeks to answer the question “How can we build computer systems that automatically improve with experience, and what are the fundamental laws that govern all learning processes?” – Carnegie Mellon University

Here are some keywords that I got from those 5 definitions: algorithm, data, prediction, and questions. Using those keywords, we can generalize five definitions into our own definition:

Machine learning is based on algorithms that learn from data to make prediction and answer questions.

In particular, we can split our definition into two parts: "use data," and "make prediction and answer question." We usually refer "use data" as training and "make prediction and answer question" as prediction. And the thing we use to train and predict called a model.

We train our model to make better and more precise prediction using a group of data called datasets. This trained model can be used to predict other unseen data.

Process of doing Machine Learning

Our key is data

You may see that the term "data" appear a lot in machine learning. You guess it! Data is the key to advance machine learning, and machine learning is also the key to unlock huge potential hidden in everyday data.

There is more

We have learned about ML definitions, why it's useful, and some of its application. This is just some in-depth concepts of Machine Learning. ML is an extremely broad field, so I will try to keep each blog as concrete as possible but still retain the main ideal and purpose of it. Next time, we will learn about each type of machine learning and the details process of doing Machine Learning to give you a better sense of which approach to use for a given data set.