Machine Learning

Arthur Samuel defined Machine Learning as a "Field of study that gives computers the ability to learn without being explicitly programmed". He made this statement in 1959 and the field has been growing every since.

Machine learning is typically broken into three classifications:

  1. Supervised learning - these algorithms are presented with sample data that includes the desired results. The algorithms are trained on a subset of the sample, then verified and adjusted using another subset and finally used to predict the results for the third subset.
  2. Unsupervised learning - the algorithm is presented with sample data wherein teh desired result is unknown. The algorithm is responsible for discovering the structure of the data and the results itself.
  3. Reinforcement learning - the algorithm interacts with a dynamic environment to determine optimal decision making. A great example of such learning is self-driving vehicles that must react to traffic conditions that are constantly changing.
Linear Regression and Logical Regression are two of the most common algorithmic models that are used to predict outcomes either for future conditions or classifying various conditions and outcomes.

Neural Networks have been used extensively for analysing problems such as character and facial recognition as well as for self-driving vehicles and robots. These are used for Deep Learning.

Contact SandOaks to learn more about how you can accelerate your business.