These are great first attempts! However, neural networks are notoriously bad at working with tabular data. You'd might be better served using a traditional ML model (e.g., linear regression, SVM).
Regardless of whether you're using a neural net or otherwise, you should normalize/transform your input features and the output feature (i.e., your closing price). ...
How to Standardize Image With ImageDataGenerator
Standardization is a data scaling technique that assumes that the distribution of the data is Gaussian and shifts the distribution of the data to have a mean of zero and a standard deviation of one.
Data with this distribution is referred to as a standard Gaussian. It can be beneficial when training neural ...
I can recommend that you always look for any dataset or kaggle problem related to the particular solution that you want to carry out and see the architectures with the best results in the evaluation metrics of these datasets.
You can also check at https://paperswithcode.com/
where they have the papers tagged by type of problem and their link to github with ...
In general yes, the predicted probability can be used in this way. However it's important to take into account that this probability is a prediction itself, i.e. the model could be wrong about it. For example the model may predict a probability of 99% positive for an instance which is actually negative. As usual, it cannot be assumed that the model is ...
This is not exactly the same accuracy, the difference between 0.58 and 0.62 might be significant.
This might be perfectly normal, there's no reason to expect the accuracy to be necessarily different for different classifiers.
Accuracy is a very simple evaluation measure for binary classification, it's suitable only if the data is perfectly balanced. It's ...
Evaluation should always be specific to the target task and preferably rely on some unseen test set.
The target task is paraphrasing, so the evaluation should be designed to check externally how good the generated sentences are as paraphrases. Usually this kind of task (the output is similar to Machine Translation) is evaluated by using a gold standard set ...
The LSA community seems to have first used the word “embedding” in Landauer
et al. (1997), in a variant of its mathematical meaning as a mapping from one space or mathematical structure to another. In LSA, the word embedding seems to have described the mapping from the space of sparse count vectors to the latent space of SVD dense vectors. Although the word ...
There are a couple of things I would suggest:
Reshape the input data: It looks to me that you want to analyse a time series if IQ-values and each time series is 128 datapoints. In this case you probably want to treat I and Q as the channels respectively and convolve over ther 128 points. To do this the input data needs to be of shape (128, 2). Right now you ...
So this isn't strictly about deep learning models or any applications to time series forecasting, but a more general approach to handle covariates below.
The link includes a python package based off three cited journal articles on how the authors using Gaussian Mixture Models to handle this issue. It looks like ...
To answer your question, I would say transformer based T5 models have performed state of the art in this particular problem statement i.e. Question Generation. Do give a try to this awesome repo: https://github.com/patil-suraj/question_generation. This guy is a regular contributor in HuggingFace.
In order to evaluate a model (see how well it works) you usually set aside one part of the data (often randomly chosen) which is NOT used for model training (so you split the data in train and test data).
You use the train data to train some model. Models are usually trained (or estimated) based on optimization of some function (the "loss"). In ...
Training error is simply an error that occurs during model training, i.e. dataset inappropriately handle during preprocessing or in feature selection. On the other hand testing errors are slightly different, such as model overfitting and underfitting, etc.
The reason for the train/validate/test distribution (which I found out in a painful way) is that you will get good results if you tweak a model to fit to a test set. It could be completely random data, but if you calculate enough features, and tweak the hyperparameters of your model, you will get a relatively high (and misleading) level of accuracy.
You are describing ensembling, combing a collection of models.
The most common ensembling design patterns that could be applied in your situation are:
Stacking - The output of model becomes the input of another model.
Bagging - Each model votes for final result.
Since this is a unsupervized problem, you need to try to extract "topics" using topic modeling. There are a number of tools available in Python, e.g. from sklearn or spacy.
Extract text from PDF
Text preprocessing (lowercase, stemming etc)
Return "topic" per page
You can test multiple different models. In fact, in industry, it's pretty common to start out with a simple model and then build a more advanced model. The first test of a simple model might be a K-NN. Then you might build something more advanced like a random forest.
For those who are interested in this question, I finally found out some useful metrics that perform well in comparing distributions (and different from KL-divergence):
Wasserstein metric, Energy metric, Shannon-Entropy metric, Maximum Mean Discrepancy metric. They are metrics in the sense that they satisfy the properties of a mathematical metric(symmetric ...
Generally, the most useful measure is the total "wall" time it takes to run the entire training script.
If training is defined as including hyperparameter search, then hyperparameters should be included. The result is the longest empirical time which is useful for estimating time it would take to reproduce similar results.
I think you have an issue with the number of elements in your dataset for analyzing a neural network with 257 features.
Consider reducing the number of features. Are all of them mandatory? What is the correlation between them? What is the mutual information between all these variables?
Consider adding more data to you dataset. Is that possible? Could you add ...
I think that the issue is related to a lack of enough data for a neural network problem (LSTM).
You mentioned you added four years, but this still might not be enough. How many data elements do you have for experiment? What is the dispersion between those elements?
My recommendation is to try to approach it with other machine learning algorithms; maybe ...
It could be due to a lack of data. Your data seems to be over a year, and many data dynamics are seasonal ones. An end of year data would not have the same shape as the rest of the year. Consequently, it would be better to train the data at least over a year (preferably 2 or 3 years to let it learn frequent patterns), and then check the model with a ...
I can't comment -- where this would be more applicable -- but your y_train is class encoded (e.g., this sample's label is class 1), which is a single output. When your data are fed into the model w/ 10 output nodes, the model doesn't know what to do considering your y_train has 1 output for each sample.
A solution would be to one-hot encode your outputs (e.g....
That is commonly called entity linking, the task of assigning a unique identity to entities. Your issue in particular is name variations, the same entity might appear with different textual representations / surface forms.
Clustering is not the most useful way of solving name variations since clustering is unsupervised.
There are many ways to approach ...
For specific models, there is some hyper-parameter optimization algorithm and some toolbox that do it.
If I say the famous one, the grid search is a well-known(but not optimum) algorithm for hyperparameter optimization. For more info you can visit this page
Moreover, you can find a good tutorial in this page1 or page2
but for cascading some architectures and ...
Here 256-d can come from one of the three things termed in the paper as A, B, C.
A → Zero Padding Projections
B → projection shortcuts are used for increasing dimensions, and other shortcuts are identity.
C → all shortcuts are projections.
(as mentioned on page 6 of the paper)
In the Below Visualization of Resnet Block, a parallel 1x1 Operation is ...
It might depend on what embedding algorithm you are referring to, but you often decide your vocabulary size in advance, and any words are given the OOV (out of vocab) or UNK (unknown) token.
Sorting your vocabulary by frequency (in a representative sample of data) is the most sensible way to decide which words make the cut, and which get discarded.
Does it ...
What do you mean by reported speech? It might be easier to help out if you could elaborate on the end goal. What are you trying to do?
I see so what you are looking to do is to translate between active and passive voice. When it comes to techniques to do this I found several options:
You can train a model using a long short-term memory (LSTM) recurrent ...
In case of image classification one of the important thing to remember while connecting Convolutional layer output to Dense layer is to make sure Flatten the output of the Convolutional layer then pass it to Dense layer. Just add following code snippets just before your 1st Dense layer containing 2048 neurons.
There is a general principle in linguistics and consequently in NLP: the meaning of a word is represented by the context of the word, i.e. the words around it.  In NLP this principle is the basis of distributional semantics, which is used in every NLP application involving semantics (almost of them).
This means that statistically the meaning of a word ...
For imblanced data I recommend using false positive ratio instead of precision. In contrast to recall, precision is affected by the positive negative ratio, false-positive ratio (or 1 minus this number) is not affected by the data distribution, only by your model.
One of the major reason for not getting accuracy in your neural network.
As this is audio data based classification more better approach is to used convolutional neural network by first performing fast fourier transformation on the data.
You can always try following style neural network architecture
norm_layer = preprocessing.Normalization()
In short, the answer is yes, your intuition is correct. The problem you are trying to solve is a Multiagent Reinforcement Learning problem and single-agent approaches don't work well here. One of the main obstacles is the non-stationary transitions of the environment's states. There are various settings (e.g. competitive and collaborative) and training ...