Machine Learning and it’s classification

What is machine learning?

Field of study that gives computer the capability to learn without being explicitly programmed. Machine Learning can be explained as automating and improving the learning process of computers based on their experiences without being actually programmed. The main focus on machine learning is about to give an output from the machine without any user interaction. The example of ML is a search engine, image recognition, speech recognition, regression, classification, predictions, etc.

How’s it’s work?

  • Data gathering
  • Data observation
  • Divide into training validation, and test, etc.

Pre requires to learn Machine learning:

  • Calculus
  • Linear algebra
  • Statistics
  • Probability
  • Graph theory
  • Programming skill

Type of Machine Learning:

  1. Supervised Learning
  2. Unsupervised Learning
  3. Semi-supervised Learning
  4. Reinforcement Learning
  1. Supervised Learning:-

    Into this, the model is getting trained on a labelled dataset now you think what is labelled dataset?
    the dataset in which all input and output are contained. While training the model, data is usually split in the ratio of 80:20. Using 80% of data are used to give training to model and the rest of 20% are used for testing. supervised learning classified as below
     1.1 classification
     1.2  Regression

     

    Example of the supervised algorithm are:-

    Linear Regression
    Nearest Neighbour
    Decision Tree
    Support Vector Machine
    Random Forest

     

  2. Unsupervised Learning:

    Into this, we do not give any target to our module while training. The only input parameter is given to the model. the machine itself find which way it can learn. Unsupervised learning classified as

     2.1 Clustering
     2.2 Association

     

    Example of the Unsupervised algorithm are:-

    K meaning clustering
    Heretical   clustering

     

  3. Semi-Supervised Learning:-

    The name suggests working between supervised and unsupervised technique. Used this technique when dealing with data which are little bit labelled and rest large portion is unlabelled.
    Unsupervised technique user to find labelled data and that data are used to feel the input to the model.

  4. Reinforcement Learning:-  

    Into this technique, find the best path is a goal or take suitable action to maximum reward in a particular situation. The main difference between supervised and reinforcement is that in reinforcement agent decides in what to do perform a given task whereas in supervised labelled data are given as inputs.

    Main Point of reinforcement:-
    Input:- The initial point
    Output:- many solutions for any particular problem
    training:- based upon input
    The model keeps continues to learn
    The best solution is decided based on max reward

    Reinforcement learning classified as:-
    4.1 Positive
    4.2 Negative


     

 

 

Leave a Reply

Your email address will not be published. Required fields are marked *