What is Machine Learning?

Machine learning (ML) is a modern software development technique and a type of artificial intelligence (AI) that enables computers to solve problems by using examples of real-world data. It allows computers to automatically learn and improve from experience without being explicitly programmed to do so.

 To start, let's talk about the relationship between artificial intelligence or AI and machine learning. AI refers to the broad capability of machines to perform activities using human-like intelligence. Machine learning, or short form, ML, is artificial intelligence. It allows computers to automatically learn and improve from experience without being explicitly programmed to do so. Using machine learning, computers can learn from data to discover patterns and make predictions. Supervised learning is a type of machine learning technique in which every training sample from the dataset has a corresponding label or output value associated with it. As a result, the algorithm learns to predict labels or output values. You can use supervised learning to do things like predict the sale price of a house or classify objects in an image. In unsupervised learning, there are no labels for the training data. The algorithm tries to learn the underlying patterns or distributions that govern the data. Remember, in supervised and unsupervised learning, models inspect data to discover patterns, then humans use the patterns learned by the model to gain new understandings or make predictions. There is another type of machine learning called reinforcement learning, which takes a different approach. Reinforcement learning is learning what actions to take in a situation to maximize reward. It is similar to how you might train your pet. If your dog does something you want it to do, you might reward it with a treat. If it does something you don't want it to do, you might correct it with a small penalty like raising your voice just a little bit. Your dog learns to do the things that get a treat and avoid doing things that get a correction. In machine learning, reinforcement learning works exactly like this. Now, let's see how machine learning help solves problems and how this differs from traditional problem-solving. In traditional problem-solving with software, a person analyses a problem and engineers a solution in code to solve that problem. For many real-world examples, this process takes a lot of time. It might even be impossible. This is because a correct solution needs to consider numerous edge cases. For example, imagine the challenging task of writing a program that can detect if a cat is poisoned in an image or not. Traditional problem-solving would require careful attention to details like varying lighting conditions, different types of cats, colors, etc. In machine learning, we have a flexible component called a model. We also have a special program called the model training algorithm to adjust the model to real-world data. The result is a trained model that can be used to predict outcomes not part of the dataset used to train it. In a way, machine learning automates some of the statistical reasoning and pattern matching that the problem-solver would traditionally do. The flexibility of the model is the key here. The machine learning field has seen rapid and recent growth. As you start on your machine learning journey, you might see related, but different definitions of the terms. This is because machine learning is a new field at the intersection of statistics, applied math, and computer science. Each of these fields might have a slightly different formal definition for the same terms.