"No one is harder on a talented person than the person themselves" - Linda Wilkinson ; "Trust your guts and don't follow the herd" ; "Validate direction not destination" ;

January 15, 2023

Time Series Notes - Features Evaluation - Anamoly Detection

Time Series Features

  • Date time features -  hour, month, and day of week for each observation. Daylight savings or not, Public holiday or not, Quarter of the year, Hour of day, Season of the year. 
  • Lag features and window features - Business day, Quarter start, Weekly frequency
  • Rolling window statistics - moving average
  • Expanding window statistics - minimum, mean, and maximum values
  • Domain-specific features - additional research into each feature and find out domain-specific information beyond what is provided in the dataset description
  • Additional factors, such as trends, seasonality, holidays, and external economic variables.

Introduction to feature engineering for time series forecasting

Top 4 Time Series Feature Engineering Lessons From Kaggle

Time Series as Features

Ref - Link



An observation that deviates so much from other observations as to arouse suspicions that it was generated by a different mechanism



  • Python Outlier Detection (PyOD)
  • Python Streaming Anomaly Detection (PySAD)
  • Python Graph Outlier Detection (PyGOD)



Ref - Link


Keep Exploring!!!

No comments: