Machine Learning explained for complete Newbie

Photo by Andy Kelly on Unsplash

Machine Learning explained for complete Newbie

Play this article


Machine learning is a buzzword that has fascinated a lot of people and has the potential to greatly change how we interact with technology, from virtual assistants, self-driving cars, and large language models like Chatgpt and Bard, to Image generation like Stability AI and Mid Journey. In this blog post, I will explain to you the concept of machine learning in the smallest possible way so that even a newbie can grasp it.

What is Machine Learning?

Machine Learning is a subset of Artificial intelligence that is focused on teaching computers to learn without being explicitly programmed to do so.

Machine Learning involves training computers to learn from data to make predictions or make decisions. For example, you can train a computer on a bunch of e-commerce data and it will learn how to predict which product has the potential to make a hit.

Types of Machine Learning

Machine Learning is a wide field that can be divided into different types..

Supervised learning:

Supervised learning is a type of machine learning that involves training a computer with data that includes its label, For example, you can train a computer on the height, weight as well as gender of a person along with their age, and have it learn and predict the age of another person given the weight, height and gender. In summary, in supervised learning the computer is trained on data containing the features as well as its corresponding output or target.

Unsupervised Learning:

In unsupervised Learning, the machine or computer is given only a set of data without the corresponding output so it learns to understand the data and group similar data together based on what it has learned. This type of machine learning algorithm is mostly used for customer segmentation, anomaly detection, recommendation systems, image and audio processing, natural language processing(NLP) etc.

Reinforcement Learning:

"Reinforcement Learning" is one of the major categories of machine learning and artificial intelligence that focuses on training intelligent agents to make sequences of decisions in an environment to maximize a cumulative reward signal. It is often used in scenarios where an agent interacts with an environment, learns from these interactions, and adapts its behavior to achieve specific goals or tasks. Here's an explanation of the key concepts and components of reinforcement learning:

  1. Agent: The learner or decision-maker that interacts with the environment. The agent makes a sequence of decisions to achieve a goal.

  2. Environment: The external system or world with which the agent interacts. The environment provides feedback to the agent based on its actions.

  3. State (s): A representation of the current situation or configuration of the environment. The state is crucial for decision-making as it encodes all relevant information.

  4. Action (a): The choices or decisions made by the agent in response to the current state. Actions are selected from a set of possible options.

  5. Policy (π): A strategy or function that defines the mapping from states to actions. It guides the agent's decision-making process.

  6. Reward (r): A numerical signal provided by the environment after each action. The reward indicates the immediate benefit or cost associated with the chosen action. The agent's objective is typically to maximize the cumulative reward over time.

  7. Trajectory or Episode: A sequence of states, actions, and rewards that occur during the interaction between the agent and the environment. A trajectory represents one complete interaction cycle.

Reinforcement learning is mostly used in various domains, including robotics, games, self-driving cars, recommendation engines etc. It is a powerful paradigm for enabling machines to learn and make decisions in complex, dynamic environments through trial and error.

Linear Regression: Linear Regression is a popular machine-learning algorithm that aims to find the relationship between a variable (target) and one or more features. It learns to achieve this by fitting a linear equation to the data. This type of algorithm falls under the supervised learning category and it is mostly used for data where their target is a continuous value, such as predicting the price of a house or predicting the salary of a person based on the number of experience etc.

Logistic Regression: Logistic regression is mostly used to learn data where their output or target is binary 0 or 1. This is mostly used to solve classification problems and then has only 2 possible outputs. This algorithm is supervised learning and it's used in cases such as predicting if an email is spam or not, checking if the rain will fall or not or predicting if the stock market will go green or red etc.

K-Means Algorithm: The k-means algorithm falls under the unsupervised learning category.
It is also a clustering method designed to partition a dataset into 'k' distinct, non-overlapping groups or clusters. These clusters are formed based on the similarity of data points, where each cluster represents data points that are more similar to each other than to those in other clusters.

Machine learning is such an interesting field for people who are excited about AI, curious about data, and eager to explore the limitless possibilities of training intelligent systems. Whether you aspire to teach machines to recognize patterns in vast datasets, create robots that can navigate complex environments, or develop algorithms that mimic human decision-making, the world of machine learning offers an incredible journey of discovery and innovation. As technology continues to advance, the opportunities to shape the future through machine learning are boundless, making it an exhilarating path for those with a passion for pushing the boundaries of what's possible.

Thank yu for reading.