Machine learning is a process for computer to automatically train models from historical data and output results. As new data is fed, the modeling algorithms automatically train statistical models to fit new data. The process keeps going so that the models are updating to produce updated outcomes.

Six components of Machine Learning 

•  Statistical Methodologies: Modern statistics methodologies are founded on theories of statistics. In machine learning, common used methods are un-supervision learning (classifying analysis such as cluster analysis), supervision learning (linear and non-linear models) and reinforcement learning (Bayesian probability, neural network).   
• Modern Computers and Information Technologies: With dramatically increasing computer speed and storage and speedy data processing and transferring, machine learning becomes feasible and affordable to all kinds of businesses. 
• Statistical Analysis Software:  The core of machine learning is to obtain estimated values by using statistical modeling and classification for large amount of data (GB is normal). It is impossible for people to perform such complex tasks. Thanks for statistical software to get large complex computation done. Popular statistical software used in machine learning are  SAS, Python, and R. 
• Efficient Computing Programs:  Efficient Computing Programs with Sophisticated Algorithms to transform data into information, knowledge and strategies according to business rules and objectives. 
• Data availability: With technology development, more and more things are recorded as data. Data size and variety are rapidly increasing. However, turning massive unstructured data into information and knowledge needs human inputs. 
• Human Intelligences and Instructions: Without understanding data and business natures, computer can not do right jobs. Human being instruction with understanding of businesses is a key to make computers to produce right estimates, discover relationships, detect patterns, and develop strategies.