https://github.com/firmai/simple-machine-learning-glossary
Simple Machine Learning and Data Science Definitions without Copyright
https://github.com/firmai/simple-machine-learning-glossary
Last synced: 3 months ago
JSON representation
Simple Machine Learning and Data Science Definitions without Copyright
- Host: GitHub
- URL: https://github.com/firmai/simple-machine-learning-glossary
- Owner: firmai
- Created: 2019-12-21T18:38:35.000Z (about 6 years ago)
- Default Branch: master
- Last Pushed: 2019-12-21T19:31:35.000Z (about 6 years ago)
- Last Synced: 2025-10-03T23:42:24.993Z (3 months ago)
- Size: 15.6 KB
- Stars: 7
- Watchers: 3
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Machine Learning and Data Science Definitions for Business (without copyright)
The purpose of these defintions are to lean towards ease of understanding and analogy as a means to quickly communicate work to management. Use it freely and don't feel the need to cite anyone.
----
**Target** is what you want to predict.
*They have decided to use the next month's revenue as the target and not the number of customers served*
---
**Features** is the columns of data you are using to predict the target.
*I have a list of features I am considering like the size of the house, the number of bedrooms and the number of bathrooms, to predict the value of the house (target)*
---
**Supervised learning** is a method to identify the relationship between data (called features) and something and that which you are trying to predict (called the target) using a machine learning model.
*She used a supervised learning model to predict whether or not the passanger will survive (the target) the Titanic by looking at their characteristics like age, gender and type of ticket (the data)*
---
**Unspervised learning** is a method where you only want to identify the structure in data instead of predicting something.
*He used a supervised learning model to predict whether or not the passanger will survive (the target) the crash by looking at their characteristics like their age and gender (the data)*
---
**Reinforcement learning** is a method for an agent to learn the best action to take in an environment given the agent's current state so as to maximise future expected rewards.
*She used a reinforcement learning model to dynamically price their website merchandise (environment), it tracks the online user's current behaviour (their state) and then changes the price of items (action) in response based on previously learned patterns to optimise sales (future rewards/value function)*