Machine Learning - An Intro

Summary
Here, I discussed about the first step to achieve the idea of 'Artificial Intelligence'. And where 'Deep Learning' falls in the world of AI. Further, I end this with introduction to 'Machine Learning'.
  • The idea of AI is to build a machine that can think like a human and do all the stuff what humans can do. Machine Learning is the first stone to throw in the well in order to build such a machine .

  • Learning ML(Machine Learning) makes you build models to predict or classify the data in order to make decisions. ML is a subset of AI.

  • Learning DL(Deep Learning) makes you build neural networks that mimics human brain. DL is a subset of ML.

So only when ML and DL are combined only then one can achieve the idea of AI.

Introduction to ML:

  • ML makes you to build models that enables machines to analyze, find patterns, classify and predict the data without being explicitly programmed.

  • There are mainly 3 types of ML. They are :

    1. Supervised ML

    2. Unsupervised ML

    3. Reinforcement ML

  • In Supervised Learning we have the training data consists of input samples along with their corresponding desired outputs. Now the idea is to train the model mapping those input and output samples to find the pattern in order to either predict or classify the new data.

  • Unlike Supervised Learning we don't have labeled data in Unsupervised Learning so the model need to generate or find the pattern by understanding the relationships within the data without any guidance.

  • Unlike supervised or unsupervised learning, reinforcement learning is based on the concept of trial and error, where the agent interacts with the data, receives feedback in the form of rewards or penalties, and learns to take actions that lead to higher rewards over time.