I have time series data of equipment readings and I am trying to predict the type of equipment depending upon the historical/live reading. I looked into Dynamic time wrap which I think might be useful but looking to find any other approach that can be used.
Can you please give us more information about the dataset you are working on? That would help to clarify your doubt in a better way.
Meanwhile, you can refer this course on time series forecasting that might help you to solve the problem:
Thanks @PulkitS. Actually I am looking to find similarity and differences between different time series in order to classify them then predicting future values.
So generally sensor has sin function( where values fluctuate constantly) while setpoint had straight line(like setting ac temp to constant value.). I am looking to classify these series into either or the category.
How many different time series do you have in your dataset? Can you please share the dataset you are working on?
For the time series forecasting , the best way to use ARIMA (Autu regression model) and Holtzwinter models in python , these two models will classify the time series data
But to increase the accuracy of this algorithm , LSTM(Long short term memory) is the best proven technique to get very accuracte accuracy.