Machine Learning Components and Applications

Study Notes

Machine learning involves four key components: data storage, abstraction, generalization, and evaluation.
Data storage is vital for holding and accessing vast amounts of data, used by humans and computers for advanced reasoning.
Different methods are used for computer data storage e.g. hard disk drives, flash memory and similar devices to store and retrieve data.
Abstraction is extracting knowledge about stored data to create general concepts.
Generalization transforms knowledge of stored data into usable information for future tasks similar to past ones.
Evaluation provides user feedback to measure the usefulness of learned knowledge, enabling improvements in the learning process.

Machine learning is a core technology in today's rapidly evolving world.
Image recognition identifies objects, persons, places, and digital images. Facebook uses this for automatic friend tagging.
Speech recognition converts voice instructions into text, used by virtual assistants (e.g., Google Assistant, Siri, Cortana, Alexa).
Traffic prediction helps users find the shortest routes and predicts traffic conditions using real-time vehicle location and historical data.
Product recommendations are used by e-commerce companies (e.g., Amazon) to suggest products based on user interest.
Self-driving cars use machine learning to detect people and objects for safe operation.
Email spam and malware filtering automatically categorizes emails as important, normal or spam.
Virtual personal assistants respond to voice commands to perform tasks (e.g. play music, make calls, schedule appointments).
Online fraud detection uses machine learning to identify fraudulent transactions by analyzing transaction data patterns.
Stock market trading anticipates market trends.
Medical diagnosis uses machine learning to identify diseases.
Automatic language translation converts text between languages; Google uses neural machine translation.