First Step Towards Machine Learning: An Introduction

Original article was published by Akashrajpuria on Artificial Intelligence on Medium


First Step Towards Machine Learning: An Introduction

We are glad to stand in the Middle of Revolution from traditional programming to Machine Learning

Yes, you heard right Machine Learning topic for today’s blog.

If you want to know what is Machine Learning and how to start with it, this blog is for you

Topics :

(1)What is Artificial Intelligence?

(2)What is Machine Learning and Traditional Programming?

(3)Different Types of Machine Learning

(5)Machine Learning Industrial Application.

What is Artificial Intelligence?

If machines can think like us, they can understand what we speak, speak in our language, can see what we see, and process it the same way then it means that they are trying to match intelligence which is specific to humans.

· Sophia is an AI robot that everybody knows, The robot who got the citizenship, can talk, tune in, and comprehend in human language.

· THIS ultimate goal of making a machine intelligent like a human, using Sensor, Machine Learning, NLP is called Artificial Intelligence.

What is Machine Learning?

Teaching a Machine the same as we teach a Human

Let us take an example and understand it better

So you have an alternative to teach a Student by case 1 or case 2

Case1: You can PROGRAM a human or cause it to learn on the off chance that you see 2 eyes and 4 legs and 1 tail: Then Print cat

Case 2: You will show an image and will name that image as a cat and let the kid make sense of the number of eyes, tail, legs.

And Definitely, we use case 2

That is the basic difference between traditional programming and Machine Learning

Traditional Programming :

· In traditional programming you hard code the behavior of the program. (ex. If {this} print {this})

Machine Learning:

· The computer does not just use a pre-written program but learns how to solve the problem itself.

A computer program that teaches the computer how to program themselves: -SO that we don’t have to explicitly describe how to perform a task we want.

So you give a cat image and a label “CAT” and the Machine will learn by itself how a cat looks.

Machine Learning and its types:

Supervised Algorithm

· The term supervised learning originates from the idea that an algorithm is learning from a training dataset, which can be thought of as the teacher.

Ex.(Training Dataset = { Image and (label: Image name) }

· In general supervised learning occurs when a system is given input and output variables with the intention of learning how they are mapped together, or related.

· Finding how Input and output variables are related, How much Output variable change if input variables are altered.

The algorithm knows the correct answer (label)– make a prediction –(if wrong) corrected (by the teacher) — improve accuracy

Once the algorithm achieves excellence teaching is stopped.

Examples of Supervised Machine Learning

(1)Given the cost of the house as label and features(area, locality, rooms) as data. You can use a Supervised algorithm to predict the cost of any house

Classification Regression

Classification: Output label is Categorical-Discrete(Fruits, Colour)

Regression: Output Variable is Continuous-Numerical(Time, Temperature)

Unsupervised Algorithm: NO LABEL

The term unsupervised learning comes from the idea that an algorithm is learning on own, No teacher-No label.

1. Unsupervised Algorithm does not know the output, Based on some pattern or similarity it clusters similar things.

2. Like in the above example unsupervised algorithm was able to cluster duck, mouse, and rabbit WITHOUT LABEL

3. Here the task of the machine is to group unsorted information according to similarities, patterns, and differences without any prior training of data.

Unsupervised learning is the training of machine using information that is neither classified nor labeled and allowing the algorithm to act on that information without guidance.

Reinforcement learning

Trains algorithms using a system of reward and punishment

· The Pacman learns by trying all the possible paths and then choosing the path which rewards him with the least hurdles

· Each right step will give the Pacman a reward and each wrong step will subtract the reward of the robot.

· The total reward will be calculated when it reaches the final reward that is completing the level.

· It is about taking suitable action to maximize reward in a particular situation

It is bound to learn from its experience.

Reinforcement learning is the training of machine learning models to make a sequence of decisions

Real-World Examples

(1)Gmail

Gmail has now got a smart reply feature which will suggest small brief responses to whatever email you’ve received based on the content that is present in the email. The smart compose option will give you suggestions like greetings, closings, or some whole sentences in between while you’re busy typing the email.

(2)Netflix

At Netflix, ML has been constantly used to improvise the recommendations and personalization problems. ML has also expanded into various other streams like content promotions, price modelling, content delivery and marketing too. The entire platform seems to run 80% through the recommendation engine.

(3)Uber

ML is a fundamental part of this tech giant. From estimating the time to determining how far your cab is from your given location, everything is driven by ML. It uses algorithms to determine all these effectively. It does these by analyzing the data from the previous trips and putting it in the present situation.

Congrats for going to the last piece of instructional exercise

Presently you can reply and comprehend the accompanying question

(1)What is Artificial Intelligence?

(2)What is Machine Learning, its types, how it works, and a few examples.

Your Task:

Comment More Examples of

(1)Supervised — Regression Algorithm

(2)Unsupervised Learning

(3)Where would we be able to utilize Reinforcement Learning?

(4)Your conclusion on beneath Image?

Thanks for understanding and starting with Machine Learning.

I have created a video for a better understanding of the topic: Introduction to Machine Learning

https://www.youtube.com/watch?v=Sf6v_ZanX38&t=809s