adagrad

a matrix of input data that created inside gradDescent.learn
function.

inputData

a matrix of output data that created inside gradDescent.learn
function.

outputData

a list of parameter that customize the learn.
       
<ul>
<li> <code>rowLength</code>: a integer of data length (row).
</li>
<li> <code>theta</code>: a matrix of float number of current model value.
</li>
<li> <code>alpha</code> : a float value for learning rate.
</li>
<li> <code>momentum</code> : a float value to give a constant speed to learning process..
</li>
<li> <code>smooth</code> : a float value to handle zero division issue in certain learning method.
</li>
<li> <code>stochastic</code> : a boolean value to enable stochastic, which mean to select one random
                   value in data train, instead process all data train. 
</li>
<li> <code>accelerate</code> : a boolean value to enable accelerate in the learning with momentum.
</li>
<li> <code>maxIter</code> : Adaptive Moment Estimation method to calculate gradient.
</li>
</ul>


list


This is internal function of learning stage that implement Adaptive Subgradient 
method to create model.


An implementation of various learning algorithms based on Gradient Descent for dealing with regression tasks.
The variants of gradient descent algorithm are :
Mini-Batch Gradient Descent (MBGD), an optimization to use training data partially to reduce the computation load.
Stochastic Gradient Descent (SGD), an optimization to use a random data in learning to reduce the computation load drastically.
Stochastic Average Gradient (SAG), a SGD-based algorithm to minimize stochastic step to average.
Momentum Gradient Descent (MGD), an optimization to speed-up gradient descent learning.
Accelerated Gradient Descent (AGD), an optimization to accelerate gradient descent learning.
Adagrad, a gradient-descent-based algorithm that accumulate previous cost to do adaptive learning.
Adadelta, a gradient-descent-based algorithm that use hessian approximation to do adaptive learning.
RMSprop, a gradient-descent-based algorithm that combine Adagrad and Adadelta adaptive learning ability.
Adam, a gradient-descent-based algorithm that mean and variance moment to do adaptive learning.

adagrad: Adaptive Subgradient learn

Description

Usage

Arguments

Value

References