Prediction Task

A prediction task is a decision task whose performance is based on approximate correctness.

• Context:
  • Input: one or more Data Records.
    • optional: Training Data.
  • output: one or more Predicted Values.
    • optional: one or more Prediction Confidence Values (including error bounds).
  • Performance Measure: Predictive Performance Measure.
  • It can be instantiated in a Prediction Act.
  • It can be solved by a Prediction System (that implements a prediction algorithm).
  • It can range from being a Nominal Value Prediction Task (such as class predictioning) to being an Ordinal Value Prediction Task to being a Numeric Value Prediction Task.
  • It can range from being a Real-Time Prediction Task to being an Offline Prediction Task.
  • It can range from being a Population-level Prediction Task to being an Item-level Prediction Task.
  • It can range from being a Model-based Prediction Task (that also requires a predictive model) to being a Model-free Prediction Task.
  • It can be an IID Prediction Task, a Sequence Prediction Task (e.g. temporal predictioning), or a Structured Prediction Task.
  • It can be a Prediction Task with Confidence Prediction, such as a prediction task with error bounds.
  • It can be a Method-Specific Prediction Task, such as: sample statistic point estimation.
  • It can range from being a Uni-Predictor Prediction Task to being a Multi-Predictor Prediction Task.
  • It can range from a Theoretical Prediction Task (such as a mathematical conjecture) to being an Applied Prediction Task.
  • …
• Example(s):
  • Forecasting Tasks, such as: weather forecasting.
  • Next Decision Prediction Tasks, such as: movie recommendations.
  • Credit Scoring.
  • a Population Prediction Task, such as direct marketing predictions for a population.
  • …
• Counter-Example(s):
  • Root Cause Analysis.
  • Disease Treatment.
• See: Supervised Learning, Decision Making, Problem Solving, Ranking Task.

---

---

References

2018

• http://theatlantic.com/technology/archive/2018/05/machine-learning-is-stuck-on-asking-why/560675/?single_page=true
  • QUOTE: .. Many of my AI colleagues are still occupied with uncertainty. There are circles of research that continue to work on diagnosis without worrying about the causal aspects of the problem. All they want is to predict well and to diagnose well. I can give you an example. All the machine-learning work that we see today is conducted in diagnostic mode— say, labeling objects as “cat” or “tiger.” They don’t care about intervention; they just want to recognize an object and to predict how it’s going to evolve in time. ...

2016

• Ajay Agrawal, Joshua Gans, Avi Goldfarb. (2016). “The Simple Economics of Machine Intelligence.” In: HBR, November 17, 2016
  • QUOTE: This matters because prediction is an input to a host of activities including transportation, agriculture, healthcare, energy manufacturing, and retail.
    When the cost of any input falls so precipitously, there are two other well-established economic implications. First, we will start using prediction to perform tasks where we previously didn’t. Second, the value of other things that complement prediction will rise.
    Lots of tasks will be reframed as prediction problems

2015

• Yann LeCun. (2015). “Augmented Knowledge: Teaching Machines to Understand Us." EMTech-2015
  • QUOTE: Prediction is the essence of intelligence.

---