45
votes
Accepted
Does gradient descent always converge to an optimum?
Gradient Descent is an algorithm which is designed to find the optimal points, but these optimal points are not necessarily global. And yes if it happens that it diverges from a local location it may ...
- 13.9k
41
votes
Accepted
Why not always use the ADAM optimization technique?
Here’s a blog post reviewing an article claiming SGD is a better generalized adapter than ADAM.
There is often a value to using more than one method (an ensemble), because every method has a weakness.
40
votes
Should a model be re-trained if new observations are available?
When new observations are available, there are three ways to retrain your model:
Online: each time a new observation is available, you use this single data point to further train your model (e.g. ...
- 501
32
votes
Accepted
Should a model be re-trained if new observations are available?
Once a model is trained and you get new data which can be used for training, you can load the previous model and train onto it. For example, you can save your model as a ...
- 2,316
31
votes
Accepted
Are there any rules for choosing the size of a mini-batch?
In On Large-Batch Training for Deep Learning: Generalization Gap and Sharp Minima there are a couple of intersting statements:
It has been observed in practice that
when using a larger batch ...
28
votes
Accepted
Is Gradient Descent central to every optimizer?
No. Gradient descent is used in optimization algorithms that use the gradient as the basis of its step movement. Adam, Adagrad, ...
- 396
27
votes
Accepted
Difference between RMSProp with momentum and Adam Optimizers
(My answer is based mostly on Adam: A Method for Stochastic Optimization (the original Adam paper) and on the implementation of rmsprop with momentum in Tensorflow (which is operator() of struct ...
- 664
23
votes
Accepted
How many features to sample using Random Forests
I think in the original paper they suggest using $\log_2(N +1$), but either way the idea is the following:
The number of randomly selected features can influence the generalization error in two ways: ...
oW_♦
- 6,264
20
votes
local minima vs saddle points in deep learning
Let me give an explanation based on multivariate calculus. If you have taken a multivariate course, you will have heard that, given a critical point (point where the gradient is zero), the condition ...
- 6,031
18
votes
Guidelines for selecting an optimizer for training neural networks
AdaGrad penalizes the learning rate too harshly for parameters which are frequently updated and gives more learning rate to sparse parameters, parameters that are not updated as frequently. In several ...
18
votes
Does gradient descent always converge to an optimum?
Asides from the points you mentioned (convergence to non-global minimums, and large step sizes possibly leading to non-convergent algorithms), "inflection ranges" might be a problem too.
Consider the ...
- 1,227
14
votes
Accepted
local minima vs saddle points in deep learning
This is simply trying to convey my intuition, i.e. no rigor. The thing with saddle points is that they are a type of optimum which combines a combination of minima and maxima. Because the number of ...
- 156
11
votes
Accepted
Can overfitting occur in Advanced Optimization algorithms?
There is no technique that will eliminate the risk of overfitting entirely. The methods you've listed are all just different ways of fitting a linear model. A linear model will have a global minimum, ...
- 4,139
10
votes
Choosing a learning rate
Copy-pasted from my masters thesis:
If the loss does not decrease for several epochs, the learning rate might be too low.
The optimization process might also be stuck in a local minimum.
Loss being ...
- 18.8k
9
votes
Accepted
Why is learning rate causing my neural network's weights to skyrocket?
You might find Chapter 8 of Deep Learning helpful. In it, the authors discuss training of neural network models. It's very intricate, so I'm not surprised you're having difficulties.
One possibility (...
- 341
9
votes
Why is my generator loss function increasing with iterations?
I think that there are several issues with your model:
First of all - Your generator's loss is not the generator's loss. You have on binary cross-entropy loss function for the discriminator, and you ...
- 2,200
8
votes
Why do we use gradients instead of residuals in Gradient Boosting?
Hmmm, I am little perplexed by your question.
In gradient boosting, we do use the residuals. The residuals are the gradients.
You can check my simple implementation of gradient boosting. This is ...
- 3,410
8
votes
Accepted
How does binary cross entropy work?
When doing logistic regression you start calculating a bunch of probabilities $p_i$ and your target is maximize the product of those probabilities (as they're considered independent events). The ...
- 251
8
votes
Accepted
Why aren't Genetic Algorithms used for optimizing neural networks?
Training Neural Networks (NNs) with Genetic Algorithms (GAs) is not only feasible, there are some niche areas where the performance is good enough to be used frequently. A good example of this is ...
- 28.6k
8
votes
Is Gradient Descent central to every optimizer?
According to the title:
No. Only specific types of optimizers are based on Gradient Descent. A straightforward counterexample is when optimization is over a discrete space where gradient is undefined.
...
- 9,247
8
votes
Accepted
How many times is backprop used in epoch?
It depends on the type of gradient descent or respectively your batch size: One epoch means that your neural net (NN) has applied the forward pass on all examples of your training data, i.e. it has "...
- 5,360
8
votes
Difference between RMSProp and Momentum?
Optimizers evolved with small Fix/Improvement on the previous one. So, if you will read in sequence, you will have a better understanding. In this context, RMSProp was a fix on Adagrad and it was an ...
- 5,544
7
votes
Is reseating passengers a reinforcement learning problem?
Reinforcement learning is more about interacting with an environment, and while this could be posed as an RL problem, I think using Global Optimization would be a more direct approach.
Essentially ...
- 2,351
7
votes
Accepted
Mathematical formulation of Support Vector Machines?
Your understandings are right.
deriving the margin to be $\frac{2}{|w|}$
we know that $w \cdot x +b = 1$
If we move from point z in $w \cdot x +b = 1$ to the $w \cdot x +b = 0$ we land in a ...
- 1,164
7
votes
Accepted
Is it possible to get worse model after optimization?
Is it possible that after running the optimization my score won't get better (and even worse?) ?
Yes, theoretically, by pure luck, it is possible that your initial guess, before optimization of hyper-...
- 1,500
6
votes
Choosing a learning rate
Learning rate , transformed as "step size" during our iteration process , has been a hot issue for years , and it will go on .
There are three options for step size in my concerning :
One is related ...
- 399
6
votes
Accepted
Simple example of genetic alg minimization
Here is a trivial example, which captures the essence of genetic algorithms more meaningfully than the polynomial you provided. The polynomial you provided is solvable via ...
- 6,798
6
votes
Guidelines for selecting an optimizer for training neural networks
My personal approach is to pick the optimizer that is newest (i.e. newest-published-in-a-peer-reviewed-journal), because they usually report results on standard datasets, or beat state of the art, or ...
- 241
6
votes
In Neural Nets, why Use Gradient Methods as Opposed to Other Metaheuristics?
It would be a waste of information; the gradient is available, so use it and save time.
There is reason to believe that the local optima are good; see, for example, Choromanska et al. (notes).
Over-...
- 10.5k
Only top scored, non community-wiki answers of a minimum length are eligible