Machine learning is a type of machine intelligence that allows computers to learn and improve. It is based on the idea that we can build systems which allow our data to do the talking by finding patterns in vast quantities of information. Depending on what problem they are trying to solve or how accurate an answer needs to be, different types of machine-learning models need to be trained. There are four different types of machine learning models.
- Supervised learning.
- Unsupervised learning.
- There is semi-supervised learning.
- Reinforcement learning.
What is supervised learning?
Supervised learning is a machine learning model training technique in which the machine learning models are trained by providing them with example inputs and outputs. For machine learning to be considered supervised, there must be some feedback mechanism which uses the result of the machine ‘s prediction to teach it how to perform better in the future. The machine classification model is told that it is wrong. Reinforcement learning has a concept of reward for right action taken. Key points for supervised machine learning
- There are labelled examples in the training dataset.
- The goal of supervised learning is to use the dataset to produce a model that takes a feature vector x as input and output information that allows deducing the label.
- Regression and classification problems are solved by supervised learning.
- li > Linear regression/li > li > Lasso/ridge regression/li > li > Decision trees/li >
- Some real-world examples of applications using supervised learning are housing price prediction, ad click prediction, image classification, machine translation, speech recognition, disease diagnosis, and face detection.
There is a picture explaining supervised learning.
What is unsupervised Learning?
Unsupervised learning is a machine learning model training technique in which machine learning models are not provided with any labelled data, and they must learn from the input/ environment themselves. Even though an example is missing some information, unsupervised machine-learning techniques try to find patterns in a pool of unlabelled examples. There are two types of learning.
- The method of clustering relies on creating clusters from the input data. The datapoints that have similarities will result in being part of the same clusters.
- The second method of learning is association, in which the predictions are made based on the rules found in the input data.
There are some key points about machine learning.
- There are unlabelled examples in the training dataset.
- There is a hidden pattern within the dataset where the output is not preset.
- Complex clustering and association problems can be solved with unsupervised learning models.
- Hierarchical clustering is one of the examples of unsupervised learning. K-means clustering. li > Principal component analysis/li > li > DBSCAN/li >
- Customer profiling, fraud detection, machine quality inspection, machine failure prediction are some real-world examples of applications using unsupervised learning.
There is a difference between supervised and supervised learning.
What is semi-supervised learning?
A machine learning task that uses a combination of labeled and unlabeled examples for training is called semi-supervised learning. Semi-supervised learning assumes the use of both labeled and unlabeled data in order to train on, but it does not assume that all of the labels need to be provided by humans Semi-supervised machine learning is a way to improve supervised machine learning by using unlabeled data.
There are important points in relation to semi-supervised learning.
- There will be a small amount of labelled data and a large amount of unlabelled data in the training data.
- The goal is to improve supervised learning by using unlabelled data.
- Problems of classification, regression, clustering and association can be solved with semi-supervised learning models.
- li > Generative models/li >, li > Discriminative models/li >, and li > Self learning machine/li > can be used for semi-supervised learning. /li >
There is a picture of learning.
What is reinforcement learning?
Machine learning is used to learn how to act in an environment in order to maximize a reward. The goal of reinforcement learning is the same as other machine learning techniques, but it does this by trying different actions and then rewards or punishes them based on their effectiveness in meeting your goals
There are three main types of machine reinforcement learning.
- Reinforcement learning can be model based.
- Policy based machine learning.
- Machine learning is value based.
The following points are related to reinforcement machine learning.
- Machine learning agents and models thrive in environments where they can see features in data.
- Each action of the agent brings different rewards and can transfer the agent to another state.
- Reinforcement learning agents are used to make the system learn a policy.
- The policy is a function of the features of a state that is considered as an input, and the outputs are an optimal action to implement in that state.
- The anticipated average reward is maximized by an action that is ideal.
- The goal of reinforcement machine learning is long term.
- Self-driving cars, machine chess playing and machine tutoring systems are examples of applications using reinforcement learning.
- Machine learning systems that represent reinforcement learning include Q-learning, policy iteration and deep Q network.
There is a picture of reinforcement learning.
There are some great mind maps on machine learning.
Machine learning is starting to solve complex problems. We talked about different types of machine learning tasks such as supervised machine learning. Machine learning techniques can be used to solve real-world problems if you have a good understanding of them. There are a lot of online courses you can take to learn more about machine learning. You can watch our channel for free online machine learning courses.