New answers tagged

0

You should munge the data until you have a target value that aligns with the goal of the project. Then fit a model. It sounds like you want the target to be a binned number of items {(1-10), (10-20), (20+)}. Each row should be unique Order_ID with one of those three target values.


1

Kaggle's NAB: a variety of sources such as AWS server metrics, Twitter volume, advertisement clicking metrics, traffic data, and more. Data is labeled. Kaggle's Wafer: manufacturing data, 2K datapoints, 143 labeled anomalies. Measures are taken every 10 milliseconds.


-2

Try boosting and random forests. You would need a long enough time series. A good reference (with examples in R) is James, G., Witten, D., Hastie, T., & Tibshirani, R. (2017). An Introduction to Statistical Learning: with Applications in R. Springer. Usefulness of particular features can be assessed via variable importance plots.


0

It sounds likes you have non-stationary data which can be a challenge to model. One option is to completely discard all older data and only train/test on newer data. This way the model will only capture the newer relationships. This approach assumes there is enough newer data to train a model. Another option is heavily regularize the model. The goal of ...


0

This is an active area of research called Human Activity Recognition. There are several public datasets available to cross-validate your methods, and you might want to start here: UCI HAR Dataset. There's a paper that accompanies the dataset that describes their preprocessing methods, so you'll want to have a look at that and see if anything helps in terms ...


0

I think these are often used colloquially as synonyms, but let's try to find the differences. Each of them begins with "Time Series" (TS). So the difference lies in the three following terms. here with my interpretation: Analysis - wanting to describe and understand characteristics the observed data coming from the generating function$^1$. ...


1

One option could be applying the Fourier transform which transforms time to the frequency domain. It is often easy to find recurring patterns when signals are represented in the frequency domain.


0

I resolved it by removing X because it's an index.


0

Try changing the plot type on Plotly to line: import plotly.express as px fig = px.line(df, x="Date", y="Close", title='COMERCIAL INTERNATIONAL BANK') fig.show()


1

Let's take a TS data = [ 1, 2, 3, 4, 5, 7, 8, 9, 10 ] Call the function with these parameters sequence_length=5, sampling_rate=1, sequence_stride=1, shuffle=False, batch_size=2 shuffle, batch_size has no role in TS data creation. It will come into effect when you iterate on the returned Dataset. In this case, we will have the following data points, [ 1, 2, ...


1

The issue was resolved once I used Tensor View to reshape the mini-batches for the features in the training and in the validation set. As a side note, view() enable fast and memory-efficient reshaping, slicing, and element-wise operations, by avoiding an explicit data copy. It turned out that in the earlier implementation torch.unsqueeze() did not reshape ...


0

Euclidean distance metric is unsuitable for time series...In short, it is invariant to time shifts, ignoring the time dimension of the data. If two time series are highly correlated, but one is shifted by even one time step, Euclidean distance would erroneously measure them as further apart. You might use it to compare your time series if you are extracting ...


0

It could be a number of different reasons but when I had that problem in the past it was usually due to too high of a learning rate or the optimizer. I recommended either dropping the initial learning rate or going with vanilla SGD. Occasionally I saw problems with Adam particularly if you have no warmup. You might want to try more general hyperparameter ...


1

The data either has duplicates or you may have some rows with empty values. I removed my rows at the bottom that had empty values separated by comma which resolved the issue


0

@user575406's solution is also fine and acceptable but in case the OP would still like to express the Distributed Lag Regression Model as a formula, then here are two ways to do it - In Method 1, I'm simply expressing the lagged variable using a pandas transformation function and in Method 2, I'm invoking a custom python function to achieve the same thing. ...


0

After implementing and training this model and understanding the details better, it seems the short answer is NO, I have to remake and train the entire model. It takes about 7 minutes to train about 3,500 series with about 100 datapoints each from a seed dataset. This may grow as the incremental ETL adds data indefinitely.


0

This answer has an implementation for audio spectrograms, but it is exactly the same for any multivariate time-series. Here is the most relevant part of the code. To not have any overlap, set window_hop = window_size. window_size = 64 window_hop = 30 start_frame = window_size end_frame = window_hop * math.floor(float(frames.shape[1]) / window_hop) for ...


1

The tough part about this problem is evaluating what moves are "correct" per se. In a fighting game sequence, there may be 2 or more moves which both work in theory in the current frame. If you are optimizing for knocking the other character out of frame, it would be useful to build a reward set which optimizes for this. It may be worthwhile to ...


0

As I tried to predict the position without normalizing... the error was in the data. After normalizing the positions everything worked as expected.


Top 50 recent answers are included