22 lines
1.2 KiB
Markdown
22 lines
1.2 KiB
Markdown
---
|
||
title: One-Shot Learning
|
||
---
|
||
|
||
# One-Shot Learning
|
||
|
||
Humans learn new concepts with very little need for repetition – e.g. a child can generalize the concept
|
||
of a “monkey” from a single picture in a book, yet our best deep learning systems need hundreds or
|
||
thousands of examples to grasp any object even upto a point of decent accuracy. This motivates the setting we are interested in: “one-shot” learning, which
|
||
consists of learning a class from a single (or very few) labelled example.
|
||
|
||
There are various approaches to One-Shot learning such as [similarity functions](https://www.coursera.org/lecture/convolutional-neural-networks/one-shot-learning-gjckG),
|
||
[Bayes' probability theorem](https://www.youtube.com/watch?v=FIjy3lV_KJU), DeepMind has come up with it's own version of Neural Networks using the One-Shot learning approach!
|
||
|
||
|
||
|
||
### More information:
|
||
* [Siraj Raval on YouTube](https://www.youtube.com/watch?v=FIjy3lV_KJU&feature=youtu.be)
|
||
* [Andrew Ng (Deeplearning.ai)](https://www.coursera.org/lecture/convolutional-neural-networks/one-shot-learning-gjckG)
|
||
* [Scholarly article](http://web.mit.edu/cocosci/Papers/Science-2015-Lake-1332-8.pdf)
|
||
* [Wikipedia](https://en.wikipedia.org/wiki/One-shot_learning)
|