Online Learning Task

An Online Learning Task is a learning task that is an online/serial decisioning task (where soon after the prediction is made, the true label of the instance is discovered).

• Context:
  • It can range from being a Supervised Online Learning Task to being an Unsupervised Online Learning Task.
  • It can be solved by an Online Learning System that applies an (online learning algorithm).
  • It can range from being a Heuristic Online Learning Task to being a Data-Driven Online Learning Task.
  • It can range from being a Reinforcement Learning Task to being a ...
  • …
• Example(s):
  • a Batch Learning Task,
  • a Follow-the-Leader (FTL) Task,
  • an Online Convex Optimization Task,
  • an Online Portfolio Optimization Task.
  • an Online Learning-to-Rank Task,
  • an Online Subgradient Descent (OSD) Task,
  • …
• Counter-Example(s):
  • an Incremental Learning Task,
  • a Lazy Learning Task,
  • an Offline Learning Task,
  • a Reinforcement Learning Task.
• See: Web-based Learning Task, Active Learning, Adversarial Game, Self-Supervised Learning Task, Exploration/Exploitation Tradeoff, Incremental Learning, Credit Assignment Task.

---

---

References

2017

• (Auer, 2017) ⇒ Peter Auer. (2017). "Online Learning". In: (Sammut & Webb, 2017).
  • QUOTE: Online learning and its variants are one of the main models of computational learning theory, complementing statistical PAC learning and related models. An online learner needs to make predictions about a sequence of instances, one after the other, and receives feedback after each prediction. The performance of the online learner is typically compared to the best predictor from a given class, often in terms of its excess loss (the regret) over the best predictor. Some of the fundamental online learning algorithms and their variants are discussed: weighted majority, follow the perturbed leader, follow the regularized leader, the perceptron algorithm, the doubling trick, bandit algorithms, and the issue of adaptive versus oblivious instance sequences. A typical performance proof of an online learning algorithm is exemplified for the perceptron algorithm.

2014

• (Wikipedia, 2014) ⇒ http://en.wikipedia.org/wiki/online_machine_learning Retrieved:2014-11-20.
  • Online machine learning is a model of induction that learns one instance at a time. The goal in on-line learning is to predict labels for instances. For example, the instances could describe the current conditions of the stock market, and an online algorithm predicts tomorrow's value of a particular stock. The key defining characteristic of on-line learning is that soon after the prediction is made, the true label of the instance is discovered.

2010

• (Bartlett, 2010) ⇒ Peter L. Bartlett. (2010). http://videolectures.net/mlss2010au_bartlett_onlinelearning/

2007

• (Blum & Monsour, 2007) ⇒ Avrim Blum, and Yishay Monsour. (2007). “Learning, Regret Minimization, and Equilibria.” xx xxx.
  • ABSTRACT: Many situations involve repeatedly making decisions in an uncertain environment: for instance, deciding what route to drive to work each day, or repeated play of a game against an opponent with an unknown strategy. ...

---