Time Series data in Machine Learning Algorithms [closed]

Closed. This question needs to be more focused. It is not currently accepting answers. Want to improve this question? Update the question so it focuses on one problem only by editing this post. Closed 5 years ago. Improve this question I am relatively new to machine learning and have recently had some exposure to Decision … Read more

CO2 and global temp: Can you model relation between two measures if one is monotonic?

CO2 and global temp: Can you model relation between two measures if one is monotonic? In 1998 the “hockey stick” paper was published in Nature (Mann Bradley Hughes Nature April 1998 v392 pp 779-787). This was supposedly mainly Mann;s dissertation work, so I will say “Mann.” Mann basically took a bunch of predictors and used … Read more

Looking ahead at seasonality in time series modeling without overfitting

In forecasting the performance of many agents in a time series, there is a strong seasonality component, in addition to non-seasonal features for each agent. How can I capture the overall seasonal trends in my model building process without overfitting? Currently, I’m using a time-series cross validation approach to fitting a Random Forest Regressor or … Read more

Predicting next event time

Problem definition: Predict user’s next event date, based on previous event occurrences. The aim is to inter-corporate time dependent and time independent features. Data: +10 year transactional data generated by millions of users. 80% of users have less than 3 events. Prediction: Next event date I’ve gone through many similar questions, the most related ones … Read more

Intuition/interpretation for the value of the spectral density calculated at the zero frequency

Is there any easy way to think about the value of the spectral density at the zero frequency? How do people interpret it usually? I have been studying some time series on my own and encountered a CLT for dependent data (covariance stationary and ergodic processes). The sufficient conditions of that CLT fail if that … Read more

Exercise on cointegrated processes

I have this time series problem that I cannot solve: Considering two I(1) independent stochastic processes zt and xt and a stationary stochastic process wt, I have the following DGP: yt=xt+zt+wt. Suppose that I misspecify the DGP and regress yt only on xt and zt. Will I be able to find cointegration between xt and … Read more

Pattern Recognition Time Series via FFT

I came across this interesting article where the author used FFT to discover some patterns in a time series. I am new to this kind of analysis and have maybe some basic questions about it. How do you compute the Frequency when getting the FFT? I used the fft function in MATLAB with some data … Read more

TimeSeriesSplit for multiple features in training set [closed]

Closed. This question is off-topic. It is not currently accepting answers. Want to improve this question? Update the question so it’s on-topic for Cross Validated. Closed 4 years ago. Improve this question I am trying to use Time-Series Split to establish a training and testing dataset and encountered the problem that I can not incorporate … Read more

Look at the time series plot, how to tell if the data is “random” and nonnormal or not?

I tried to solve this exercise in the book of “Time Series Analysis with Application in R”: Simulate a completely random process of length 48 with independent, chi-square distributed values, each with 2 degrees of freedom. Display the time series plot. Does it look “random” and nonnormal? Repeat this exercise several times with a new … Read more

OLS – Correlation of Residuals Over Longer Time Frames

I’m running a standard financial time series analysis – calculating multi-variate OLS against 6 different factors on a monthly returns oriented time series. The time series seem to be pretty well behaved, with 1 or 2 lags of autocorrelation and some skewness. In looking at the monthly residuals (e.g. excess returns / alphas), they are … Read more