I am looking for some advice for my problem. I have a spike detection algorithm that detections outliers in a signal. I have annotations in my dataset that tells me if I am in 3 events, basal, event A and event.

I made an heuristic decision model based on the time between the spikes and the amplitude of the spikes but I want to explore using another model and training with my annotations.

I have very basic ML knowledge. I have trained some models but in other problems I have sequential data in windows ( I was getting features in a windowed appraoch) however, in this case, I am not using windows but the spikes are detected when they occur.

How would you recommend addressing this problem? Is it possible to train a ML model that from several characteristics such as inter-spike time, spike amplitudes, or any other feature?



You must log in to answer this question.

Browse other questions tagged .