4
votes
Accepted
What is the "semantic level"?
In language theory, there are generally several admitted levels that can be studied in relation with one another or independently. The semantic level is the one dealing with the meaning of the text ("...
- 166
4
votes
Accepted
Which unsupervised learning technique can be used for anomaly detection in a time series?
So if I understood correctly:
You have data from 2 sensors in time: Ar flow and BackGas Flow (SCCM, what is that?)
You have that data for multiple products.
1 - Since it is relatively low dimensional, ...
3
votes
Accepted
Why is it harder to achieve good results using neural network based algorithms for multi step time series forecasting?
ANNs & RNNs can be used to create some great models in many different domains, including time-series forecasting. However, across all of these domains, they suffer from the problem of hyper-...
- 9,147
3
votes
Can CNNs be applied to non-image data, given that the convolution and pooling operations are mainly applied to imagery?
You can use CNN for time-series data. The Convolutional Recurrent Neural Network (RCNN) is one of the examples.
Convolutional layers basically extract features from images. It is not related to time-...
- 71
3
votes
Accepted
Can CNNs be applied to non-image data, given that the convolution and pooling operations are mainly applied to imagery?
Usually, you need to ensure that your convolutions are causal, meaning that there is no information leakage from the future into the past. You could start by looking at this paper, which compares ...
- 1,108
2
votes
Can non-sequential deep learning models outperform sequential models in time series forecasting?
You are right CNN based models can outperform RNN. You can take a look at this paper where they compared different RNN models with TCN (temporal convolutional networks) on different sequence modeling ...
- 1,108
2
votes
Accepted
Feature extraction timeseries, model compatibility
there's a lot to un-pack in this question.
Why do they only pick 500 rows?
my guess: in order to keep the example running quickly. tsfresh usually takes a while to calculate its features. note that ...
- 211
2
votes
Why is it harder to achieve good results using neural network based algorithms for multi step time series forecasting?
Given the (usual) higher architectural complexity of ML models compared to more classical forecasting models, ML models might also require more data, otherwise they might just overfit the training ...
- 39.1k
2
votes
Do we need non-linear activation function in neural networks whose task isn't classification?
Yes, definitely.
In the simplest example, predicting an output value for a time series is classification. You take in the previous time steps and classify what is the most likely next value. You could ...
- 1,346
2
votes
Accepted
Inner working of Bidirectional RNNs
The second implementation looks more correct and inline with how Bidirectional is defined. Specifically, bidirectionality doen't change the forward/backward logic of either direction, and just merges (...
- 276
2
votes
How to predict time series with accuracy?
Things like this a really hot topic in research right now, and it's very difficult to get high accuracy on a chaotic system like the stock market. That being said, I would probably recommend ...
2
votes
How to deal with the time delay in reinforcement learning?
Most RL algorithms assume a discretization of time (although RL can also be applied to continuous-time problems [1]), i.e., in theory, it doesn't really matter what the actual time between consecutive ...
- 39.1k
2
votes
How to deal with the time delay in reinforcement learning?
This problem has been formally termed as Delayed MDP (Katsikopoulos & Engelbrecht, 2003)[1] - the actions generated are not instantly applied to the environment and/or the captured observations ...
- 531
2
votes
Accepted
Rescaling time-series data with very spiky pattern for training data in LSTM network
First, if your data has a minimum of 0 and maximum of 5000, 1000 will get rescaled to .2 and 5000 will get rescaled to 1. So it's not a .001 difference as you suggest.
If you just used a regular loss ...
- 1,171
2
votes
Is there a way to parallelise the RL training on multiple stocks to avoid the memory issue?
Take a look at:
Deep Reinforcement Learning for Automated Stock Trading where the 30 Dow Jones stocks are trained using OpenAI Gym.
The code is here and the published paper is here.
An excerpt from ...
- 1,853
2
votes
Accepted
Transformer model is very slow and doesn't predict well
Edit
I just noticed that the model you are referring to is built very differently than the transformer from Attention is All You Need since it only uses one half of the architecture. Thus my answer ...
- 476
2
votes
Accepted
In a Temporal Convolutional Network, how is the receptive field different from the input size?
As you can see in Fig. 2 of the WaveNet paper the receptive field is 5, but the input size is larger (16). The receptive field defines what a single output neuron can see (see arrows in Fig. 2).
The ...
- 208
2
votes
Should we always use the usual no leakage train-val-test splt in time series?
Most time series contain time-dependent information. That means that the time series has order. The temporal information, therefore, cannot be randomly sampled without losing some pertinent ...
- 1,853
2
votes
Are there public examples of AI models that predicted short-term price well?
AFAIK, unfortunately no.
However, we can get a feel of these beasts from some clues:
They do machine learning. E.g. see this competition of Two Sigma on Kaggle, and this competition by Optiver. The ...
- 313
2
votes
Accepted
Which preprocessing is the correct way to forecast time-series data using LSTM?
A standard method for pre-processing time series data for neural network architectures, such as an LSTM, is to normalize the data. Good tutorials will include this step. There are several variations ...
- 1,853
1
vote
Accepted
What's the best architecture for time series prediction with a long dataset?
Yes, LSTM are ideal for this. For even stronger representational capacity, make your LSTM's multi-layered.
Using 1-dimensional convolutions in a CNN is a common way to exctract information from time ...
- 261
1
vote
Predicting a day's data
Just for clarification: your description (1 sample per minute) does not match the example data (far fewer data points which is understandable, but also two data points in one minute which contradicts ...
1
vote
What would be a typical pre-processing and data normalization pipeline for time series data (for non-linear models such as neural networks)?
I just finished my masters thesis project for multivariate time series prediction.
The standard approach is to normalize the data using minmax or z-score (there is one research paper that found it
...
- 127
1
vote
What would be a typical pre-processing and data normalization pipeline for time series data (for non-linear models such as neural networks)?
I would say any normalization such as min-max or standard deviation is fine as far as the scaling factor is provided as a feature, since time-series of different scale might behave differently.
- 86
1
vote
Accepted
Which type of feature extractor do you suggest to classify sensor data?
There could be multiple possible ways to extract the features. One would be to use RNNs for a temporal relationship as the input data is time-series.
1
vote
Recommendations or resources for neural network/deep learning for time series application?
I'm currently working with Temporal Convolution Networks (TCNs) for making predictions with time series data (link to article here: https://medium.com/@raushan2807/temporal-convolutional-networks-...
- 191
1
vote
How do I test an LSTM-based reinforcement learning model using any Atari games in OpenAI gym?
If I understand your problem correctly, you can test on just about any environment, and just omit parts of the observations to ensure your RNN is learning. For example, you can test on cartpole, ...
- 1,106
1
vote
How to exclude sections of bad data from time-series data before training an LSTM network
Your idea is a good one.
Another idea is to upsample or aggregate your data. For example, average by week if you generally have a couple of missing days in every week.
A similar question on Stack ...
- 1,853
1
vote
What are modern state-of-the-art solutions in prediction of time-series?
An interesting model I encountered in a course is Facebook Prophet. Prophet takes into account trends, seasonality, and holidays for its predictions. As you can probably guess, this is a model that ...
Only top scored, non community-wiki answers of a minimum length are eligible