Скачать или смотреть Machine Learning Tutorial | Part-1 Model-Based Approach for Forecasting | Rohit Ghosh | GreyAtom

Prediction is the second part of dealing with time-series data, the first being analysis. Analysis is breaking the data into its components: patterns, trends and noise. Prediction comes after that is done, as it name suggests, it takes historical data and predicts the future based on that.

Time-series prediction is very similar to other prediction approaches in machine learning, with data-specific nuances. The major difference is that, with other data, there is no value that can be used to predict the value of a target variable; whereas with time-series data, historical values can be used to predict future values of target variables.

The main focus of this discussion about time-series data is the feature engineering that is possible. The first approach we look at is the model-based approach, which identifies the target variable, the dependent variables and the independent variables. The techniques used in classification and regression problems in the past will reappear here too.

To get notified about our latest content, subscribe now to our YouTube channel: https://www.youtube.com/channel/UCtha...

To stay updated with the latest trends, visit:
Facebook:   / greyatomschool  
​​Twitter:   / greyatomschool  
​​LinkedIn:   / grey.  .
Instagram: https://www.instagram.com/greyatom_sc...
Website: https://greyatom.com