Machine Learning-based Application

A Machine Learning-based Application is a data-driven application that involves the significant use of machine learning pipelines (which applies ML algorithms to solve ML tasks).

• Context:
  • It can (often) be associated to an ML Use Case.
  • It can range from being a Classification Application to being a Ranking Application to being an Estimation Application.
  • …
• Example(s):
  • a Netflix's Movie Recommendation Service (which continually improves by feedback from recommendation feedback).
  • a Google's Language Translation Service (which continually improves by new ML techniques)
  • an NLP-based Application, such as an NLG application.
  • a Voice Agent Application (which likely uses ML techniques).
  • …
• Counter-Example(s):
  • a Expert System-based Application.
• See: ML Tool, Automated Decisioning.

---

---

References

2017

• (Wikipedia, 2017) ⇒ https://en.wikipedia.org/wiki/machine_learning#Applications Retrieved:2017-6-19.
  • Applications for machine learning include:
    • Adaptive websites; ...; Bioinformatics; Brain-machine interfaces; Cheminformatics; Classifying DNA sequences; Computer vision, including object recognition;
    • Detecting credit card fraud; Information retrieval; Internet fraud detection; Machine learning control; Medical diagnosis.
    • Natural language understanding; Optimization and metaheuristic; Online advertising; Recommender systems; Robot locomotion; Search engines.
    • Sentiment analysis (or opinion mining); Sequence mining; Speech and handwriting recognition.
    • Financial market analysis; Structural health monitoring; Syntactic pattern recognition; User behavior analytics

2014

• Laura D. Hamilton. (2014). “Six Novel Machine Learning Applications.” In: Forbes.
  • QUOTE: 1) Automating Employee Access Control; 2) Protecting Animals; 3) Predicting Emergency Room Wait Times; 4) Identifying Heart Failure ...; 5) Predicting Strokes and Seizures ...; 6) Predicting Hospital Readmissions ...

2013

• Laura D. Hamilton. (2013). “10 Surprising Machine Learning Applications.” In: www.lauradhamilton.com July 30, 2013
  1. Seal Mobile ID is trying to recognize the user of a mobile device based on accelerometer data (how he holds and moves the phone).
  2. The Online Privacy Foundation sponsored a competition to see if it's possible to predict whether someone is a psychopath based on his twitter usage. (According to the leaderboard, you kind of can.)
  3. Fast Iron wants to predict the auction sale price of a piece of heavy equipment — essentially create a Blue Book for bulldozers.
  4. Similarly, Carvana is building a model to determine if a car bought at auction is a lemon.
  5. Marinexplore and Cornell University are trying to identify whales in the ocean based on audio recordings so that ships can avoid hitting them.
  6. Dunnhumby and hack/reduce are trying to predict in advance whether a product launch will be successful or not.
  7. Oregon State University is looking to determine which bird species is/are on a given audio recording collected in field conditions.
  8. Amazon is looking for a model to "predict an employee's access needs, given his/her job role." If new employees are starting with inadequate permissions, then it is a costly time suck for them to submit access request paperwork, get supervisor approval, and get granted access by IT. If Amazon is able to create a smarter permissions system with a machine learning model, they can save quite a bit of time and money.
  9. Benchmark Solutions is trying to predict the trade price of U.S. corporate bonds.
  10. StackOverflow wants a model that will predict which new questions will be closed.

2012

• (Domingos, 2012) ⇒ Pedro Domingos. (2012). “A Few Useful Things to Know About Machine Learning.” In: Communications of the ACM Journal, 55(10). doi:10.1145/2347736.2347755
  • QUOTE: … developing successful machine learning applications requires a substantial amount of “black art” that is difficult to find in textbooks.

2007

• (Bishop & Lasserre, 2007) ⇒ Christopher M. Bishop, and Julia Lasserre. (2007). “Generative Or Discriminative? Getting the Best of Both Worlds.” In: Bayesian Statistics, 8.
  • QUOTE: In many applications of machine learning the goal is to take a vector [math]\displaystyle{ \bf{x} }[/math] of input features and to assign it to one of a number of alternative classes labelled by a vector [math]\displaystyle{ \bf{c} }[/math] (for instance, if we have C classes, then [math]\displaystyle{ \bf{c} }[/math] might be a C-dimensional binary vector in which all elements are zero except the one corresponding to the class).

---