Machine Learning
What is Machine Learning?
Machine learning is a subfield of artificial intelligence. Its focus is on building and powering machine systems with the abilities to learn without being explicitly programmed to do so. Machine learning is nowadays everywhere, from chatbots to diagnosing medical conditions.
Machine learning works through the creation and usage of machine learning models. These models produce final outcomes with the help of data, algorithms, and other optimisation strategies.
ML can be divided into three main categories: supervised, learning, unsupervised learning, and semi-supervised learning. Supervised learning trains models with pre-defined (labeled data) input and output data in order to predict future outcomes. Unsupervised learning trains models that can analyse, group, and find hidden patterns within an unknown (unlabeled) data. Semi-supervised learning is a modelling technique that uses both supervised and unsupervised learning.
Reinforcement learning is an additional area of machine learning not included in the above three categories. Reinforcement learning does not use any data (labeled or unlabeled). Instead, RL operates through trial and error, rewards, environment, and intelligence agents. An agent(s) is placed in an environment, the agent takes an action, and the agent is rewarded accordingly (whether positively or negatively). The focus is to maximise the agent’s rewards.
Applications of Machine Learning
The following are some of the numerous applications of machine learning out there.
Social Media | 1. Identifying fake news 2. Recommending friends and content 3. Filtering spam messages |
Banking | 1. Detecting fraudulent transactions 2. Personalising customers’ needs 3. Classifying documents |
Advertising | 1. Segmenting customers’ characteristics 2. Predictive targeting and testing 3. Personalised ad targeting |
Healthcare | 1. Identifying diseases and diagnosis 2. Personalising medicine 3. Discovering pharmaceutical medicine |
Recommendation systems | 1. Recommending products 2. Market segmentation 3. Ranking systems |
Optimisation | 1. Resources allocation 2. Real-time data streaming 3. Production improvement |
Search engines | 1. Detecting patterns 2. Natural language processing 3. Imagine classification and searching |
Telecommunication | 1. Detecting of illegal accessing 2. Identifying fake callers 3. Speech recognition |
Traffic control | 1. Object detection 2. Object classification 3. Recommendation systems for traffic congestions |
Self-driving cars | 1. Identifying objects 2. Predicting pedestrians and other vehicles’ behaviour 3. Interpreting complex vision tasks |
Stock market trading | 1. High-frequency trading 2. Finding patterns in data 3. Assessing risks |
Robotics | 1. Training movements and improving accuracy 2. Image detection through computer vision 3. Natural language processing |
Email detection systems | 1. Identifying spam emails 2. Clustering emails 3. Email automation |
Next: ML Life Cycle