Understanding Machine Learning in One Article

Understanding Machine Learning in One Article

by
Understanding Machine Learning in One Article
This article is reproduced from the public account sigua’s inner voice, a frontline small developer in the mathematics department, updating original articles on data structures and algorithms | deep learning | workplace technologies.This public account is followed by Google TensorFlow programmers, Apple company programmers, Microsoft programmers, a contestant from the ultimate quiz show, CTOs of tech companies, and provincial champions. Everyone is welcome to follow.
This article is approximately 6000 words, recommended reading time 10+ minutes
This article provides a detailed interpretation of ensemble learning in model algorithms and its application in deep learning in a visual format.
Table of Contents
  • Overview of Machine Learning
    • Learning from Data
    • What is Machine Learning
  • Types of Machine Learning
    • Supervised Learning
    • Unsupervised Learning
    • Reinforcement Learning
  • The Process of Machine Learning
  • How to Measure the Quality of a Model
    • Overfitting
    • Training, Testing, and Validation Sets
    • The Confusion Matrix
    • Accuracy Metrics
  • Basic Data Analysis
  • Overfitting and Underfitting
  • Bias and Variance

Overview of Machine Learning

  • Learning from Data

Before diving deeper into this topic, let’s take a step back and think about what real learning is. A key concept we need to consider for machines is learning from data. Humans and other animals can adjust our behavior by learning from experiences. Learning gives us flexibility in life; we can adapt to new situations and learn new skills. A significant part of human learning is remembering, adapting, and generalizing: recognizing that the last time we were in this situation (saw this data), we tried a specific action (produced this output), and it worked (correctly), so we will try it again, and if it doesn’t work, we will try something else. In summary, the last statement is about recognizing similarities between different situations so that something applied in one place can be used in another. That’s why learning is useful; we can apply our knowledge in many different places.

  • What is Machine Learning

Thus, machine learning is about enabling computers to modify or adjust actions to make them more accurate, with accuracy measured by the similarity between the chosen action and the correct action.
In essence, it is about getting computers to learn from data.
Formal Definition:
Machine Learning is about building systems that can learn from data. Learning means getting better at some task, given some performance measure.

Types of Machine Learning

We loosely define learning as getting better through practice on a task. This leads to two important questions: how does the computer know if it is getting better, and how does it know how to improve? There are several different possible answers to these questions, leading to different types of machine learning. Now, we will consider whether the machine is learning: we can tell the algorithm the correct answer to the problem (labeled), so it can solve it correctly next time; we want to tell it some correct answers so that it can

Leave a Comment