What is Machine Learning:
AS humans, we read,observe and listen to learn any information to solve/perform a problem/task.
Machines need to be told, how to do any task either by programming or by hard coded instructions.
If we have sets of data, that can be used to train a machine to perform the required task.
So machines can also be learnt by data sets.This concept is called machine learning.
Process of Machine Learning:
1.Get the data
2.Label the data
3.Train the model
4.Evaluate with the data
5.Use the model
What are different techniques for machine learning:
1.Supervised Learning.
It needs labeled data.Data need to be assigned with labels before proceeding to machine learning algorithms.
Classification,Regression techniques are designed to work on supervised learning.
2.Unsupervised Learning.
It does not need labeled data.It needs data to be categoriezed into groups to solve the ML problem.Clustering algorithms are designed to work on
unsupervised Learning.
There are different algortihms,which implement the above techniques.
1.Linear Regression
2.Logistic Regression
3.Decision Trees
4.K-Means
5.Naive Bayes
So in any any ML problem, we need to extract the features of data to define the model.
Deep Learning:
Deep Learning is another technique of machine learning, which solves the highly complex ML problems like pattern recognition
by modeling a neural network of inputs.
Do we need to implement ML algorithms:
Standard Algorithms are already implemented by advanced APIs like Tensorflow, AWS ML API.It is required to learn,how to use those algorithms.
But sometimes,it may be required to implement our own algorithm to solve the specific real world problem.
AS humans, we read,observe and listen to learn any information to solve/perform a problem/task.
Machines need to be told, how to do any task either by programming or by hard coded instructions.
If we have sets of data, that can be used to train a machine to perform the required task.
So machines can also be learnt by data sets.This concept is called machine learning.
 |
Process of Machine Learning:
1.Get the data
2.Label the data
3.Train the model
4.Evaluate with the data
5.Use the model
What are different techniques for machine learning:
1.Supervised Learning.
It needs labeled data.Data need to be assigned with labels before proceeding to machine learning algorithms.
Classification,Regression techniques are designed to work on supervised learning.
2.Unsupervised Learning.
It does not need labeled data.It needs data to be categoriezed into groups to solve the ML problem.Clustering algorithms are designed to work on
unsupervised Learning.
There are different algortihms,which implement the above techniques.
1.Linear Regression
2.Logistic Regression
3.Decision Trees
4.K-Means
5.Naive Bayes
So in any any ML problem, we need to extract the features of data to define the model.
Deep Learning:
Deep Learning is another technique of machine learning, which solves the highly complex ML problems like pattern recognition
by modeling a neural network of inputs.
Do we need to implement ML algorithms:
Standard Algorithms are already implemented by advanced APIs like Tensorflow, AWS ML API.It is required to learn,how to use those algorithms.
But sometimes,it may be required to implement our own algorithm to solve the specific real world problem.