New answers tagged

1

The way LSTMs are capable of learning long term dependencies is by keeping a cell state which serves as a memory of sorts. This cell state is updated based on the values of different gates within the cell, forget gate being one of them. The forget gate looks at the previous hidden state and the current input and outputs a number between 0 and 1 for each ...


1

Welcome to the community! Well ... the answer needs a bit of explanation. It is not about MLPclassfier or another python function. It is about background in Signal Processing/Analysis specially for bio-signal. For different bio-signals like EEG there has been feature extraction studies for long while. For example in your problem, BCI, you probably know ...


0

It's because 80,000 samples is probably too much training data. Usually, statistical forecasting models are not trained on the entire training set like ML models. Try reducing your training data significantly, and incorporate only a certain number of seasons as your training data. Example: If your data is hourly, one season can be 24 samples. Then, the ...


0

I agree, the simple version of this problem isn't really a time series problem. There are some ways to deal with irregular intervals, but, houses are not all the same and we do not have a series of sale prices for it in most cases. What may of course be very predictive is time. Housing prices probably have a long-term trend that's linear over the space of a ...


0

I think the first question you should ask before you start going off on a deep learning model is, can you tell when the failure is going to occur just by looking at a plot of your data? If you can't, then no model will help you deduce when a failure will occur. You shouldn't overlook some basic models also such as exponential or poisson distribution models ...


0

At first sight it looks like you can't really use the time, since you have two completely disconnected series. But you can still try to see whether the physiological variables for one worker help predict whether an accident happens (or maybe how many accidents happen). The main problem with that is that you only have 14 workers. In any case I would start ...


1

Try these two approaches: First: Make a model, using any ML algorithm and divide your data into train and test. Now using the previous features, check the train and test accuracy. Now add the new features to the previous ones, again divide data into train and test. Check the train and test accuracy. If the new features help improve the test accuracy ...


0

I would remove the sites with more than a threshold (say60%) na data (for eg site 32). For the other sites with na data, i would fill the na data columns with average values. This way i would be able to use the entire dataset for training. While testing, use cross-validation to ensure a good accuracy. Sine you did not mention what methods have you already ...


0

Adam's answer does seem to make the most sense, however, I am not sure about the second statement "Polluting sequential data with non-sequential information". So recently I trained a character-level LSTM model, in which I just appended a non-sequential feature in the end of the sequential features. The model learned how to differentiate that pretty well. ...


0

It depends on what information you want to capture : If you want to capture the passing of time, encoding your date as days since a reference date might be a good idea. If you want to encode the cyclicality of time (months in a year), you can encode your month variables on a circle.


1

Here are two ways to model the problem. The first one is simpler, the second one is more complex but closer to your original statement of the problem. Store as an input feature You can consider the store as a feature to pass to your LSTM. With two different stores, just add a binary input feature "store" where store A is 0 and store B is 1, for example. ...


1

According to your definition (consecutive, order matters, max +/-2 difference), it's not a fuzzy matching case. It's just a minor variant of searching a subsequence: for i=0 to len(source)-len(test) { j=0 while (j<len(test)) && (abs(source[i+j]-test[j]) <= 2) { j++ if (j == len(test)) { // match found } } This is the simple ...


0

You can encode time series to images using image encoding methods like Recurrence Plots (RP), Gramian Angular Field (GAF), Markov Transition Field (MTF). See the following article: Estebsari, A.; Rajabi, R. Single Residential Load Forecasting Using Deep Learning and Image Encoding Techniques. Electronics 2020, 9, 68. https://doi.org/10.3390/...


0

Check out the shap library. I think that could help you. https://github.com/slundberg/shap


0

Yes, you can make a model, but the number of days until departure is only one feature that determines the price. I agree with Erwan that time of the year plays a role. I would say the occupancy rate (how many seats are booked divided by maximum seats) is probably even more important. If you think about this problem from the perspective of the airline, if ...


0

You certainly need to add at least one other variable representing the time of year, because from your graph it's clear that the fare can't be predicted accurately using only the time until departure: for the same day you can have many points representing different fares. That makes sense, since the fares are going to be very different depending if the ...


1

I will go through your question one by one: How to use time series for this data You can train an RNN multivariate regressor, by feeding time series of your variables. Your first layer would be recurrent (LSTM or GRU), and provided with the following input_shape: ( batch size , input size , Number of variables ) I have to only two dimension bike ...


Top 50 recent answers are included