Machine Learning Basics

What is Machine Learning ? 

Basics of Machine learning is an application of artificial intelligence (AI) that provides the ability to automatically learn and improve from experience without explicitly programming the system. Machine learning focuses on the development of computer programs that can access data and learn to use it for themselves.

The learning process begins with observation or data, such as examples, direct experiences, or instructions, so that patterns can be discovered in the data and in the future to make better decisions based on the examples we provide. The primary objective is to allow computers to learn automatically without human intervention or assistance and to adjust tasks accordingly.

Imagine, you have been planning to learn guitar for several days. But due to this epidemic situation you are not able to go out and take offline classes. Therefore, you decided to learn online from YouTube. You open your YouTube application and search for guitar tutorials and you have found some tutorials on it. You watched some videos and selected one to learn from.

Have you ever wondered how this actually works, how do you find related videos?

Yes, you are right. Its machine learning.

YouTube uses machine learning algorithms to give you the best results. One of the types of machine learning algorithms is - recommendation system.

This is only an example! There are many others. Amazon Echo understands you and can answer your questions. Your vacuum cleaner can also make its way around your home while Netflix is ​​recommending videos that match your profile! Apple can recognize your friend's face in the photo you took. Machine learning has become a big part of our daily lives, and it is not going anywhere anytime soon.

Therefore, in this tutorial today, we will try to learn the basics of machine learning and how it actually works.

So, let’s understand what is machine learning?

Machine learning is nothing else but getting a computer to program itself. If there is programming automation, machine learning is automating the process of automation.

Writing software is a bottleneck, we don't have enough good developers. Let data work instead of people. Machine learning is the way to make programming scalable.

As you know something about machine learning. But really, you might be wondering why we need this.

In traditional programming we provide data and write a program that runs on a computer and produces output.

But in machine learning we provide data and outputs that run on a computer to create a program or function. This program / function can be used in traditional programming.

Machine learning is like farming or gardening. Seeds are algorithms, nutrients are data, gardeners are you, and plants are programs.

Let us understand this with a simple example. Suppose we want to predict the value of a new house based on the size of the house, the number of bathrooms, and the number of its rooms.

We can try to create a traditional algorithm that answers this question. This algorithm would have to take into account the characteristics of the house, returning an estimated price based on three more explicit rules here. In this example, the exact house pricing formula has to be clearly known and coded. But in practice, this formula is often not known.

On the other hand, we can construct a machine learning algorithm. First, such an algorithm would define a model that could be an incomplete formula constructed from our limited knowledge. Then, the model will be adjusted by training on examples of given housing prices. In doing so, we combine a model with some data.

In general, machine learning is incredibly useful for difficult tasks when we have incomplete information or information that is too complex to be coded by hand. In these cases, we can give the information available to us for our model and give this one the "missing" information to "learn". The algorithm will then use statistical techniques to extract missing knowledge from the data.

As you understand what machine learning is. Let us try to understand what are the different types of machine learning?

Machine learning can be broadly classified into three types.

1)    Supervised Learning

2)    Unsupervised Learning

3)    Reinforcement Learning

Supervised Learning

In supervised learning, we aim to construct a model to predict the label of data based on their given characteristics. To learn the mapping between features and labels, the model has to be fitted to the given examples of features with their respective labels. We state that "the model is trained on a labeled dataset." And this is known as supervised machine learning.

Now, our predicted labels can be numbers or ranges. For example, we can construct a model that predicts the price of a house, which means that we want to predict a label that is a number. In this case, we called it as regression model.

On the other hand, we want to define a model that predicts a category, such as "dog" or "not dog", which is based on the given characteristics. In this situation, the model is called as a classification model.

Some other examples of regression and classification are:

Unsupervised Learning

Unpublished learning, we want to construct a model, which can approximate a function to describe the structure hidden from the unlabeled data. Here, we will have only input data (X) and no corresponding output variable.

The goal of the model is to discover the underlying structure or distribution in the data to learn more about the data.

For example, unhelpful learning algorithms can help answer questions such as "Are there groups between my data?" Or "Is there a way to simplify the description of my data?"

Unheard education can be broadly divided into two types:

1) clustering

2) Association

The model can look for different types of underlying structures in the data. If it tries to find clusters between data, then we will talk about a clustering model. An example of a clustering model would be a model that divides a company's customers based on their profile.

On the other hand, when you want to search for rules that describe large parts of your data, such as those who buy X, also buy Y is known as the association problem.

Some of the real use cases of the association problem are market basket analysis and web use mining and intrusion detection.

Reinforcement learning

Reinforcement learning is a type of dynamic programming, which uses algorithms to reward and punish systems.

A reinforcement learning algorithm, or agent, learns by interacting with its environment. The agent receives rewards by performing correctly and punishes for performing incorrectly. The agent learns without interference from a human by maximizing his reward and minimizing the penalty.

Some examples of reinforcement learning are

Self driving car

Aircraft control and robot speed control etc.

By now I think you understand what machine learning and its variants are.

Let me summarize for you how the machine learning model works.

Imagine, the input is given to the machine learning model which then gives the output according to the applied algorithm. If the output is correct, we take the output as the end result, otherwise we provide feedback to the training model and ask it to predict it until it learns.

I hope you have understood machine learning and this is the type of supervised, unheard and reinforcement learning. Let's see some of its applications.

Applications of Machine Learning

Some applications of machine learning are:

Web Search: Google uses this method to rank the page you are most likely to click on.

YouTube uses a recommender system to suggest you some videos based on your search history.

 Gmail uses a spam classifier to classify mail as spam or not spam.

 Uber uses a machine learning algorithm to show you an approximate price based on your given destination.

So, machine learning has many more applications in our day to day life. But do you ever wonder how it is even possible.

This is possible due to the very large amount of data that we are producing on a daily basis that is used for analysis and is also advanced in available computation power with the advancement of new technology.

The basics of machine learning from me. I hope you enjoyed this tutorial. Keep watching…

Want to know more : Insideaiml