6. Glossary
Transfer learning
A method of reusing what has been learned on one problem as an aid to solving a slightly different one. This is commonly practiced with deep convolutional networks, by supplementing the lower layers of a pre-learned network with new layers, then performing learning iterations on a new learning base.
Supervised learning
A category of statistical learning whose aim is to empirically model an input-output relationship, based on pairs of examples (xn, yn) where xn is an input vector, and yn an associated output vector (called the desired output).
Characteristic; feature
In statistical learning, a quantity calculated from the raw inputs of the examples, and which will be the input of the learned model.
Classification; classification
One of the two subtypes of supervised learning problems, corresponding...
You do not have access to this resource.
Exclusive to subscribers. 97% yet to be discovered!
Already subscribed?
Log in!
Bibliography
-
(1) - RUMELHART (D.E.), HINTON (G.E.), WILLIAMS (R.J.) - Learning representations by back-propagating errors. - Nature, 323 (6088) : 533-536 (1986).
-
(2) - CYBENKO (G.) - Approximation by superpositions of a sigmoidal function. - Mathematics of Control, Signals, and Systems, 2(4), 303-314 (1989).
- ...
Also in our database
-
CANU (S.) - Machines à noyaux pour l'apprentissage statistique. - [TE 5 255],...
Software tools
For "classic" statistical learning algorithms, the richest and most widely used software tool (containing implementations of most models and algorithms) is :
Sci-Kit Learn (Python library), http://scikit-learn.org
For deep learning of convolutional networks, the main libraries used (all of which integrate transparent use of GPUs on computers that have...
You do not have access to this resource.
Exclusive to subscribers. 97% yet to be discovered!
Already subscribed?
Log in!