New answers tagged

0

Try Kaggle, they offer 30 hours of TPU per week which might help if you are working with tensorflow or torch (TPUs have 128GB of memory). What you should try when running into memory issues is to use half precision floats which reduces memory requirements. tensorflow torch


1

Here are a few options: Commercial cloud provider: AWS is the go-to cloud provider for virtually any need, they have options for everything. Check out their options for Deep Learning. You could also buy your own hardware Finally from a data science perspective there is also the option of modifying the parameters of the model and/or data so that the training ...


2

Any exploration function that ensures the behaviour policy covers all possible actions will work in theory with Q learning. By covers I mean that there is a non-zero probability of selecting each action in each state. This is required so that all estimates will converge on true action values given enough time. As a result, there are many ways to construct ...


1

The most general and robust way is to double the training set using the "mirror image" of each sample. Also it is sound approach from a statistical learning perspective. There can be some approaches to symmetry where one hacks the NN architecture (eg by symmetrising the network) but these are usualy very fragile and do not generalise well, while ...


1

Expanding my comments into an answer. Ridge regression is a by definition an augmentation of Least Squares method, especialy for problems where the data may be highly correlated (there is what is called multi-colinearity). Lets assume the dependent variable is $y$ and $x_i$ are the independent variables. Then assume the true mapping between $X$ and $y$ is: $$...


2

This would not be possible since the two variables you are trying to predict are of a different type. You are first predicting the default label, which would be yes/no, so this is a classification problem. The second variable you are trying to predict is the prepayment percentage, which is a continuous variable, this is therefore a regression problem. You ...


1

I'm not aware of any literature specific to the case of classification based on binary features since it's just a subset of the general case, but it's definitely possible. A very common example is traditional text classification, where the document is represented as a bag of words: there are different options but each word in the vocabulary can be ...


0

A Siamese network (a network with multiple outputs) will work for such a case.


1

Saying that accuracy is measured to get how accurate the model performs, and F1 is how well the model performs This doesn't mean anything, it's obviously too vague. The first things to check in order to understand this relationship are the definitions of accuracy and F1-score. Wikipedia has a good page which explains how different classification evaluation ...


0

That's to help indicate the volume of points in that region, which would be lost due to overlap if all plotted along the horizontal line. This kind of plot is referred to as a "bee-swarm", and is somewhat similar to violin plots, strip plots, or scatter plots with jitter.


1

See this article for a bit more detail on how to better explain EFB. Here is a brief visual explanation from there with my own edits. I hope you can appreciate the high production quality of my updated graphic... To answer your main question see "Part 1 of EFB". This explains that features are ordered by their sparsity and mixed in with all other ...


1

It would depend on the approach you want to take... Based on the information you gave, I could imagine turning the data into a classification problem, whereby you cluster in the feature space of the various customer profile features. You could train the classifier on a "did buy"/"didn't buy" column. The descriptive statistics you ...


0

Both options would work. It would depend on which method you want to use, what you want to do afterwards, or computing capacity. E.g. developing a model for each city might allow you to contrast and compare the outcomes more simply. But you can also do multivariate predictions to predict for each city even once you have combined your dataset. Also if your ...


0

This appears like a standard multi class classification problem. The features are frequency counts of the basic messages leading up to the fatal message. You would have as many features as there are basic categories. The labels would be the fatal message categories themselves.


0

You can use seller mean buying price, std buying price, max buying price, min buying price, median buying price PLUS include recently user buying power to suggest the totally new product to the user given the current data that's best I can recommend although extensive data can lead to better suggestions.


Top 50 recent answers are included