Bottom Page

Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
 Keras: Time series classification

I am designing a Neural Network for a classification of time series. I have 4 classes of functions and around 50000 samples for each class. The functions have a length of about 500 points normalized so that max(abs(f(t)))=1. The 4 classes are roughly speaking 'continuous', 'discontinuous', 'in-between' and 'trash'.

What Neural Network design would you use for this particular problem?

The design I'm using currently is a CNN.

model = Sequential()
model.add(Conv1D(16, 8,input_shape=(500,1)))
model.add(Conv1D(8, 5))
model.add(Conv1D(8, 3))
model.add(Dense(4, activation='softmax'))
model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])
I'm open to all suggestions. The Network is not performing very well (85% accuracy).

Thank you for your help!

Top Page

Possibly Related Threads...
Thread Author Replies Views Last Post
  Keras. ModuleNotFoundError: No module named 'error' agatte 1 613 Sep-18-2019, 12:27 PM
Last Post: shivani
  creating new time series based on giving mean, standard deviation and skewness Staph 1 329 Aug-06-2019, 10:41 PM
Last Post: scidam
  create 10 yearly blocks from time series using pandas Staph 1 217 Jul-23-2019, 12:01 PM
Last Post: Malt
  selecting customized seasons from monthly time series Staph 3 320 Jul-14-2019, 09:40 AM
Last Post: scidam
  Keras + Matplotlib causing crash spearced 2 490 May-17-2019, 09:47 AM
Last Post: spearced
  Keras Target Problem inco 1 443 May-05-2019, 10:47 AM
Last Post: scidam
  [pandas]How to liner fit time series data and get linear fit equation and r square Sri 5 631 Apr-04-2019, 12:00 PM
Last Post: Sri
  keras tensor mainpulation output to next input JohnyBeGood 0 429 Nov-21-2018, 03:30 PM
Last Post: JohnyBeGood
  Keras: tweets classicifcation kiton 0 423 Nov-14-2018, 03:19 PM
Last Post: kiton
  Calculating median value from time data series mkaru 1 602 Aug-22-2018, 08:41 AM
Last Post: Mekire

Forum Jump:

Users browsing this thread: 1 Guest(s)