Electricity Production FORECASTING with Neural Hierarchical Interpolation ⚡
HTML-код
- Опубликовано: 7 авг 2024
- In this video, we are going to forecast the electricity production using renewable energy resources for three months. Since the time series we have is an hourly time series, this task could be referred to as long term forecasting. The model we will learn in this video is Neural Hierarchical Interpolation for Time Series Forecasting (NHITS).
Link to the data: www.kaggle.com/datasets/stefa...
The code that I can show you right now is the one that I show in the video. Please note that in the video, I run the codes live. Therefore, the error value might be slightly different for every run. Link to the code: www.kaggle.com/code/leessteph...
00:00 Intro
NHITS Theory (I will describe how the model works throughout the explanation of the three features in NHITS)
01:24 NHiTS
03:24 Feature 1: multi-rate sampling
08:07 Feature 2: doubly residual stacking
10:53 Expressiveness ratio
13:29 Feature 3: hierarchical interpolation
Application
20:15 Data, preprocessing, and TimeSeriesDataSet
33:37 Train an NHITS model
41:40 Predict the electricity production
#timeseries #deeplearning #forecast #machinelearning #renewableenergy #electricity #pytorch - Наука
Oh god were are you I wait for your videos thank you ❤❤❤
I try to use pykx package from kdb pls can you help me