Survey on Automated Machine Learning (AutoML) and Meta learning
Ashwini Doke, Madhava Gaikwad
Abstract
Automated Machine Learning is an area of research that has gained lots of research in the past few years. To build a high qualitymodel for Machine learning we need technical experts who have good knowledge in exploring various machines learning algorithm as well as to tune the hyperparameters efficiently. The size of data is increasing, it is observed that data scientist cannot address this challenging tasks due to lack of expertise and experience in the respective domain. So, there is a need to automate such crucial task inthe domain of machine learning. Metalearning is “learning to learn” like human expertise. This paper is initial survey of ongoing research in field of Meta Learning and AutoML.
Topics & Concepts
Machine learningArtificial intelligenceComputer scienceHyperparameterDomain (mathematical analysis)Meta learning (computer science)Field (mathematics)Task (project management)Hyper-heuristicRobot learningEngineeringMobile robotMathematical analysisSystems engineeringPure mathematicsMathematicsRobotMachine Learning and Data ClassificationAnomaly Detection Techniques and ApplicationsData Stream Mining Techniques