umbertogriffo / predictive-maintenance-using-lstm Goto Github PK

The test set has 100 ids and should have 100 predicted values, but why only 94?

Hi, man

You have made a pretty predictive model with LSTM. I am a Ph.D student and I want to know if your model has been published as a journal paper. If it has been published as a journal paper, I wish you could tell me the DOI of the paper or the information about the paper. Thank you very much!!

HI
im a student from a collage in Norway and im approching pyton and machine learing but my teacher isn't so good, so i try to learn it by myself (i'm a super beginner) and i have some questions:
1)in the training data we have simulated sensor data and when the engine breaks so we use it for the LSTM learning right?
2) i understood that we use traning data to see if our lstm works but why we have alredy RUL of testing data?
3) i dont know what represents the two blue and green graphs (one results in regression model and in one in binary classification) with actual data compared to predicted data
thank you and happy holidays and sorry if some questions are stupid
tom

Hi man,

You have very good explain how to use neuronal network for predictive maintenance case.
So i have a question about data. How can you create cycle feature in your dataset (train)?

Thanks

I know that the "cycle" is the time unit. And 21 sensor readings is the features. Now I have a question:

if the ground truth data is not the "cycle", the ground truth data is the sensor of s21(for example). How to utilize LSTM to predict the ground truth data(s21)? If I want to predict the values of s21 from 23 cycle to 30 cycle(sequence time series), how to predict?

such as:

Tks!