FlowTemporal's profile picture. Flow Forecast is an open source PyTorch based deep learning for time series forecasting, classification, and anomaly detection framework.

Flow Temporal

@FlowTemporal

Flow Forecast is an open source PyTorch based deep learning for time series forecasting, classification, and anomaly detection framework.

Another episode of ForecastingFridays starting in 20 minutes. Today we will be running additional experiments with the Informer model as well as fixing up the GaussianLoss function. youtu.be/s_WOaT0BI2k #DataScience #TimeSeries #MachineLearning #DeepLearning

FlowTemporal's tweet card. Forecasting Fridays

youtube.com

YouTube

Forecasting Fridays


Live coding tonight starting at 5pm EDT. We will do a step by step tutorial of using Flow Forecast to forecast solar energy youtu.be/ZcHO-k3dkX0 #deeplearning #timeseries #datascience #livecoding

FlowTemporal's tweet card. Tutorial: Forecasting Solar Energy Generation

youtube.com

YouTube

Tutorial: Forecasting Solar Energy Generation


New @weights_biases report on forecasting the flow of the Virgin River and robustly evaluating models with SHAP. Looking at SHAP plots you can more easily spot models that are overfitting test data. #DeepLearning #DataScience #MachineLearning #TimeSeries rb.gy/gklar9


#DataScientists in industry why do you still prefer to use simple time series forecasting models instead of deep learning? #DeepLearning #TimeSeries #MachineLearning #DataScience

FlowTemporal's tweet image. #DataScientists  in industry why do you still prefer to use simple time series forecasting models instead of deep learning? #DeepLearning #TimeSeries #MachineLearning #DataScience

Also stay tuned we have some upcoming cool new infrastructure innovations that will allow you to complete hyper-parameter sweeps much faster! #DeepLearning #GCP #Python #Terraform


Here is a good example notebook of forecasting multiple targets at once. In this notebook we #forecast both #COVID19 cases and deaths at once for Miami Dade county. #DeepLearning #TimeSeries #DataScience bit.ly/3jiadpN


👕Get your Flow Forecast t-shirts now. Proceeds help us cover cloud computing and other framework maintainance costs. bonfire.com/flow-forecast-…


We are experiencing some problems with tests on CircleCI timing out. As a result you may see build is failing on our repo. This is an issue with CircleCI and not our repo. Tests are passing locally. Keep on forecasting!


If you are interested in using #MachineLearning to forecast river and stream flows and how they are/will be affected by #ClimateChange check out FlowDB dataset on Kaggle. We have hourly flow data spanning 21 years (2000 to present). #hydrology kaggle.com/isaacmg/flowdb…


A lot of #DataScientists in industry automatically chose to use shallow time series models solely because they think they are simpler. However, that might not always be the best option. #DataScience #TimeSeries #MachineLearning #DeepLearning

FlowTemporal's tweet image. A lot of #DataScientists in industry automatically chose to use shallow time series models solely because they think they are simpler. However, that might not always be the best option.
#DataScience #TimeSeries #MachineLearning  #DeepLearning

United States Trends

Loading...

Something went wrong.


Something went wrong.