Passion for CS

Wednesday 13 December 2017

Introduction to Machine Learning

CS Passionate December 13, 2017 AI and ML , 0 Comments

Hello My Friends,
How are you all?

This time I am here with a new topic and want to give you some more exiting thing...
Now a day we hear a word Machine Learning majority of times, even when we read any article related to AI. So, I thought this time let me represent Machine Learning in front of you.
Actually, Machine Learning is derived from Artificial Intelligence. We can say like Machine Learning is a child of Artificial Intelligence. Machine Learning means machine can be learning without program. i.e. Machine has to learn.

Now we have a question, Why Machine need to learn?
As a human being, we need to learn something regularly to update and upgrade ourselves, same as machine also has to learn to be upgraded. Because as technology grooves, Machine becomes older, so if machine can learn on it's own, it will be beneficial for it.

Applications for Machine Learning
There are lots of Applications available for Machine Learning such as, Bioinformatics, Search Engines, Speech Recognition, Software Engineering, Game Playing etc.

How to Work with Machine Learning?

First of all for machine learning there are lots of algorithms available for different modules and even we have lots of programming languages such as, Java, Python, R etc. we can work for machine learning using these languages. But the thing is How? Because there are lots of programming languages available and we have already known these different languages but with Machine Learning? So, for Machine Learning there are lots of libraries or modules or library packages available using these libraries we can work with any language for Machine Learning Algorithm.

So, if you can work with Java, you have different types of libraries available such as, DL4J, Weka, Massive Online Analysis (MOA), Mallet etc.
Weka
Weka is GUI library for java. and generally used for data mining and analysis. it's strength lies in classification of data, clustering, association rule etc.
Massive Online Analysis(MOA)
MOA is used for data mining like weka. even we can combine these both weka and MOA. MOA is used for classification, regression, clustering etc.
DeepLearning4J
This library is specially designed for Deep Learning. It is used for commercial purpose and it is open source library. It is used for different patterns.It also works for Scala and it is super power for deep neural network and reinforcement.
Mallet
It is another Java based open source library for natural language processing. and it also supports lots of different types of algorithms. such as, Naive Bayes, Decision Tree etc.

If you are a master in Python, you have Tensorflow, Scikit - learn, Pylearn2, NuPic like modules, using these modules you can work with Machine Learning.

Tensorflow

This module is used for high level neural network. It helps you to work with CPU or GPU. It mainly written in C++.

SciKit – learn

This python module is generally used to analyze and mine data. It builds by using NumPy, SciPy and matplotlib.

Pylearn2

This module is used for neural network. But it also work with other libraries i.e. we can combine other libraries with Pylearn2.

NuPic

This module does not work only for ML algorithms but it has some other functionality. It works with hierarchical temporal memory (HTM) is available in neural network. So, we can see it's powerful side.

R is also used with Machine Larning, we have lots of libraries available such as, e1017, rpart, nnet, tree etc.

e1071

This package is used for fuzzy clustering, support vector machine, shortest path problem etc types of ML algorithms.

rpart

This package is used for recursive partitioning and regression trees type of machine learning algorithms.

nnet

This library package is used for Neural Network and log linear models in ML.

tree

As the name suggest, this library package is used for classification tree and regression tree.

Even we have lots of other languages, using these languages we can work with Machine Learning such scala, clojure etc. even C++ like languages also support Machine Learning.

This is just a brief introduction of Machine Learning, and how to work with Machine Learning in Java and other languages. such as, Python and R. using different libraries and modules.
In next article of Machine Learning, I will be there with some new information related to ML.
Till then, Friends enjoy with Machine Learning.

Thursday 30 November 2017

Introduction to AI

CS Passionate November 30, 2017 AI and ML , 0 Comments

Hello everyone,

This time I am going to represent AI ( Artificial Intelligence ). Because now a days AI is
most demanding thing for current technology and trends. And big companies such as,
Google, Facebook, Microsoft etc. accept AI and take the big advantage from AI. AI is a
real future in the technological field. AI is very broad topic. We all are aware with AI, it is related to our daily life. But we never think about it.
Now, the thing is
What we think about, Artificial Intelligence?
Robots?

Fig 1. Robot

No, it is not just enough for robots or robotics, Robotics is different thing than AI. Yes
we can say robotics is a part of AI. But it is not fully AI.

AI is machine intelligence i.e. A Machine that represents it’s intelligence and it is
made by Human. That’s why Artificial Intelligence. Machine represents intelligence
like human but human beings and animals and other creatures have natural intelligence.
And machines have artificial intelligence. AI works on predefined tasks that is defined by
men when they designed it and set some algorithms. So, machine works as per it’s
algorithms and programming.

Research Areas for AI
Today, AI is hot topic for research. Different area for research in AI is, Robotics, Natural
Language Processing, Fuzzy Logic, Neural Network, etc.

At present, all the above mentioned topics are very hot and popular today for research in
AI.
In these above domain, Robotics and machine learning both are hot favorite for
researchers.
Robotics is touching the sky i.e. it is available in all necessary areas in the
world at present, because now a days some country use robots in different areas such as,
robot for medical area, robot for business areas like automobile, finance, education etc.,
robot as home appliances etc. As we all know about sophia a humanoid. It gets the all
necessary privilege as human being.
And Machine Learning is also a part of AI. It is the learning strategies for machine
through different algorithms.
In Natural Language Processing we have lots of natural languages to communicate but this time human communicate with computer. So, we need natural language processing and in which there are different applications such as, Speech recognition etc.
Neural Network as human we have brain and through our nervous system we can send a message to our brain, same as in AI neural network works as our nervous system.
This is the brief introduction of Artificial Intelligence. In next Article I will tell you more about AI.

Scope of AI
Lots of scope for AI in current technology world. I mean at present in technology field lots of scope for AI. as I describe here Robotics, Machine Learning, Deep Learning, Speech recognition, Game Development, Computer Vision, Expert system etc. are the different scopes of AI.

What should you learn to work with AI?
To work with AI, you should learn Algorithms, Data structure, and some programming languages such as, C, LISP, Prolog, Now a days as per the current trends C++, Java or Python etc. Even Maths, specially Discrete Maths or you can say Maths with Logic, Science such as Biology, Physics etc. this mentioned things are necessary for learning AI.

Now, How to learn AI or from where to learn AI?
Some Institutes such as,