Why we need to know about Machine Learning? (ML001)
This is the first post in my Machine learning series. We will covering a lot of important portions in Machine learning series which will help you to know everything about ML.
As Andrew Ng said, "AI is the new electricity", it's time that we start knowing about Machine Learning. In this article, I will be covering the basics of Machine learning and why it has become so popular nowadays, so let's start.
What is Machine Learning?
Machine learning is method in which we make the machines intelligent. It is a study of computer algorithms. Machine learning algorithms are mathematical models which predicts or make desicions based on sample data, without being explicitly programmed to do so.
The sample data on which the model make prediction is called the 'training data'. Machine learning comes in handy in places where it is very tough to create conventional algorithms like detecting faces, email filtering.
Let's understand machine learning using a simple definition -
Suppose I have a dataset of houses in California and I want to predict what will the price of a house. The dataset looks like this -
Source - Trifork
So in this dataset you can see that we have several columns named floor space, rooms, lot size, apartment, row house, corner house and detached. These are known as features which helps the machine learning model learn the distribution well. Here our task is to predict the price of a house, so price column will be our target variable or label.
The machine learning model learns the mapping function where features are mapped to the label. Now with more number of data points, the machine learning model will be able to generalize better and upon giving unknown data it can then be able to predict the price of a house.
So, lets assume our model was trained with 5000 data points or rows of data(training in machine learning is a iterative process in which all the data point is fed into the model and with each data, the model improve itself and finally we can use it for testing).
Why Machine learning?
Machine learning has been there for a long time but why it has become so effective in the last decade, is because of two things -
1) Data - In the past we didn't have much data with which we can run these machine learning which performs the best when given a large amount of data. Nowadays, we have abundance of data which makes it easy for us to run these model.
2) Computation power - In the past we had very low computation power which made it very difficult to run these memory and compute intensive models. But nowadays, we have great computation powers in form of GPUs and compute intensive CPUs.
These two reasons has caused huge increase in ML research, ML R&D work and ML startups. As an AI engineer myself, it's the perfect time to get started with Machine learning and AI.
Machine Learning usecases -
According to me, we have just got started in Machine learning and we have a long way to go. Machine learning can be used in various fields and we will discuss a few of them so that you can get an idea.
In 2020, we use Machine learning in sectors like health, finance, agriculture, manufacturing, aerospace, mining, Information technology, military, automotive, retail, hospitality sector and others.
Machine learning is great at predictive analysis which makes it very valuable tool for these industries. We will later discuss in details about how Machine learning can be applied to these sectors.
Future of Machine learning?
With abundance of data and computational power, I find no reason why Machine learning is not gonna boom. We will see a lot of advancements in ML in the upcoming years.
In coming years we will see Machine learning disrupting the industries. From self-driving cars to facial recognition, we will see a lot of things getting automated. This brings us to the great debate - "With rise and advancement in Machine learning, will people lose jobs?" which is another topic in itelf.
In upcoming years, we will see a lot of jobs in Machine learning and a rise in Machine learning engineers. So, this is the perfect time to get started on Machine learning.