Machine Learning Introduction

Machine Learning! Sounds awesome right? Yeah a new era of creating algorithm which is data driven and self learning has been started already.So why to waste time lets explore few basic things about machine learning.

What is Machine Learning?

The field of machine learning is concerned with the question of how to construct computer programs that automatically improve with experience. Various successful machine learning application has been developed in recent years ranging from data-mining program that learn to detect credit card fraud to information filtering field where it detects person's interest and provides suggestions.
What is Machine Learning Algorithm?
A machine learning algorithm is an algorithm that is able to learn from data.
According to Michell(1997) the formal definition of Machine Learning is as follows:
"A computer program is said to learn from experience E with respect to some class of tasks T and performance measure P, if its performance at tasks in T, as measured by P, improves with experience E."
Sounds complicated? well it took few hours for me to understand this definition!. No worries we will concentrate more towards the practical point of view.

Where Machine Learning is applicable?
Many kinds of tasks can be solved with machine learning.Based on the tasks it solves there are different types of machine learning algorithms.Lets have a look into few of them

  • Classification : Imagine the situation of categorizing the news from various source into different topics like sport, film politics etc. This can be solved using classification algorithm in machine learning another example is object recognition, where the input is an images(usually described as a set of pixel brightness values), and the output is a numeric code identifying the object in the image.Modern object recognition is accomplished by deep learning(Which is really deep!). Object recognition is the main reason for getting request to tag people while uploading pics in Facebook or any other social network(Time to kill the inventor right?).           So basically In this type of task, the computer program is asked to specify which of k categories some input belongs to.
  • Regression : Wow this is something amazing type of machine learning algorithm,which if used or studied properly makes you millionaire!. Have you ever thought of predicting the share price of particular company or predicting what might be the price of the flat that you are dreaming about. Machine learning has the solution for you. In this type of task/algorithm, the computer is asked to predict a value based on the past data. Where algorithm comes up with an equation kind of expression which will calculate the value based on various factors it depends,which is given as input. 
  • Transcription : Speech recognition is one of the greatest technology now a days. Google Now and iPhone siri are leading application in speech recognition. The back boon of this technology is machine learning and deep learning! where the computer program is provided an audio waveform and emits a sequence of characters or word ID codes describing the words that were spoken in the audio recording.
  • Structured Output : Structured output tasks involve any task where the output is a vector with important relationships between the different elements. For example pixel wise segmentation of images, where the computer program assigns every pixel in an image to a specific category. Machine learning can be used to annotate the location of roads in aerial photographs,the computer program observes an image and outputs a natural language sentence describing the image.These tasks are called structured output tasks because the program must output several values that are tightly inter-related.For example the words produced by an image captioning program must form a valid sentence.
  • Anomaly detection : This is one of the main beauty of machine learning, where banks can understand you credit card transaction habits and model them so that any misuse or unexpected behavior of transaction can be identified.If a thief steals your card, the thief's purchases will often come from a different probability distribution over purchase types than your own.The credit card company can prevent fraud by placing a hold on account as soon as that card has been identified as misuse. No only credit card transactions there might be some online transactions such as online shopping etc. As we progress towards digital economy these kind of security risks are more so there is a high demand of technology which can prevent fraud activities. So basically in this type of task, the computer program will go through a set of events or objects,and flags and identifies if some of them being unusual or atypical.
  • Denoising  : Imagine a debut program in News Channels where there is chance that people's talk will get overlapped which is unpleasant.If you want to listen to some specific person without any noise machine learning has its solution.Where it filters the voice signals of  a specific person/source of sound.In this type of task, the machine learning algorithm is given with corrupted example obtained by an unknown process to obtain the clean example
There are still more types of applications which exits and became day to day part of our life.We might be using it in various ways. The world will be completely data driven, where data plays a major role in our life.Ranging from suggesting which  food you can have when you enter into the restaurant to self driving cars.

Finally these algorithms gets better and better and more approximate as the time passes as it experiences more and more data. So future of machine learning is really bright and good place to start your carrier. Will discuss few algorithms which are used in Machine Learning in detail in coming post :)

Happy Coding! 





Comments

Post a Comment

Popular posts from this blog

Introduction to Deep Learning and Artificial Neural network.

Image
Deep Learning is one of the major branch of Artificial Intelligence, where models will be trained in a different way as compared to machine learning. Before getting into what is Deep Learning and how the deep learning models are trained,I will discuss about one of the question which I faced most of the times " What is the difference between Machine Learning and Deep Learning? " In the previous post about Machine Learning, we have seen that data should have a set of features which uniquely defines the labels in case of ML and  models will be trained with the help of few ML algorithms.But consider the example of categorizing animal pictures into different classes like cats,dogs etc.if we go with machine learning approach we have to first define the feature set attributes which will classify the animal type,features like shape,eye color, size of nose etc.The drawback of using ML approach for these kind of problem is huge size of feature set. So in order to deal with such ...
Read more

Machine Learning - Work Flow

Image
Work flow usually represent the steps to be carried to obtain a desired  output. We can say its a representation of the various  stages data should go through in order to obtain a machine learning model. The out come of the Machine learning workflow is the black box model which represent the relationship between the input data and and the output information. There are 2 types of machine learning problems Supervised and Unsupervised .Now we can concentrate on supervised learning where data used will have labels along with features whereas in unsupervised learning data will be consist of only features. What do you mean by "Feature set"?  Feature set is the set of measurable properties of data that is being observed to get the result. Consider an example of judging whether a given fruit is apple or orange we need to study/observe various characteristics of given fruit such as color,taste shape etc. All the property which describes a particular object can't be...
Read more

Functional Programming in JAVA

JAVA 8 release is the package of all surprises, functional programming is one among them. Let's explore a bit on functional programming style in java... Functional programming is the natural successor to Object Oriented programming.The objective of the design of the FPL is to mimic mathematical functions to the greatest extent possible. Will start with examples for functional programming in java Normally if we want to write any functions say addition of two numbers we were following below approach: public int add(int x,int y) {    return x+y; } The above piece of code can be written in FP style like Function<Integer,Integer> adder = (x,y)->x+y; Where Function belongs to the package  java.util.function General prototype     java.util.function.Function<T,R>      where T is input datatype, R is return data type. Java has one more Interface called  BinaryOperator under  java.util.function  which ...
Read more