Machine Learning

What is machine learning?

“Machine learning is a field of study that gives computers the ability to learn without being explicitly programmed.” – Arthur Samuel (1959)

“An application of Artificial Intelligence(AI) that uses statistical techniques that provides computer systems the ability to automatically learn and improve from experience without being programmed explicitly is well known as Machine Learning.”

“Machine Learning at its most basic is the practice of using algorithms to parse data, learn from it, and then make a determination or prediction about something in the world.” – Nvidia

“Machine learning algorithms can figure out how to perform important tasks by generalizing from examples.” – University of Washington

“The field of Machine Learning seeks to answer the question “How can we build computer systems that automatically improve with experience, and what are the fundamental laws that govern all learning processes?” – Carnegie Mellon University

The more precise definition of machine learning suggested by Tom Mitchell (1998) Carnegie Mellon University, a well posed Learning Problem is a follows,

“A computer program is said to learn from experience E with respect to some task T and some performance measure P, if its performance on T, as measured by P, improves with experience E.”

Machine Learning Methods

Machine learning algorithms are often categorized as follows:

1. Supervised Learning Algorithms
2. Unsupervised Learning  Algorithms
3. Semi-supervised machine learning Algorithms
4. Reinforcement Learning Algorithms

Supervised Machine Learning Algorithms:

“Supervised Machine Learning is the task of inferring a function from labeled training data”. A supervised learning algorithm analyzes the training data and produces an inferred function, which can be used for mapping new examples. This algorithm consist of a target variable which is to be predicted from a given set of predictors. Using these set of variables, a function that map inputs to desired outputs is generated.

Methods in Supervised Learning:

• Classification
• Regression
• Forecasting

Unsupervised Machine Learning  Algorithms:

“Unsupervised Machine Learning is a task of trying to find hidden structure or pattern in an unlabeled data.” The target or outcome variable to predict or estimate is not available in this algorithm. The input variables to the algorithm are not labeled.

Methods in Unsupervised Learning:

• Clustering
• Dimension Reduction

Semi-Supervised Machine Learning Algorithms:

“The inputs to the algorithm are both labeled and unlabeled examples in order to generate an appropriate function or classifier.” Semi-supervised learning is similar to supervised learning, but instead uses combined labeled and unlabeled data. Labeled data is essential information that has meaningful tags so that the algorithm can understand the data, whilst unlabeled data lacks that information.

Reinforcement Learning Algorithms:

“The algorithm which itself learns a policy of how to act given
an observation of the world using feedback from the environment.” Every action has some impact in the environment, and the environment provides feedback that guides the learning algorithm. Using this algorithm, the machine is trained to make specific decisions. The machine is exposed to an environment where it trains itself continually using trial and error. Machine learns from past experience and tries to capture the best possible knowledge to make accurate decisions.

Applications of Machine Learning:

1. Recommender Systems
   • Content based
   • Collaborative filtering methods
2. Email Spam Filtering
3. Pricing Optimization
4. Sentiment Analysis
5. Predictions while Commuting
6. Personalized Treatment Medication
7. Videos Surveillance
8. Social Media Services
9. Chat-bot
10. Online Customer Support
11. Search Engine Result Refining
12. Virtual Personal Assistants
13. Online Fraud Detection

Summary:

Machine Learning is a  technique of mastering computer systems to learn from past experience and evolve as predictor to predict and determine present and future problems in the world. Supervised Learning, Unsupervised Learning, Semi-supervised Learning, Reinforcement Learning are the four types of Machine Learning Algorithms. Few important application of machine learning are Recommender System, Virtual Personal Assistant, Spam Filtering & Prediction while Commuting.