✳flâneur — a map of the web's best reading

ML Blog - Everything you need to know about Time Series Forecasting

mrtunguyen.github.io · 8,093 words · saved by 1 readers

Time series exist in a variety of fields from meteorology to finance, econometrics, and marketing. By recording data and analyzing it, we can study time series to analyze industrial processes or track business metrics, such as sales or engagement. Also, with large amounts of data available, data scientists can apply their expertise to techniques for time series forecasting. We will first learn how to make simple forecasts that will serve as benchmarks for more complex models. Then we will use two statistical learning techniques, the moving average model and the autoregressive model, to make forecasts. These will serve as the foundation for the more complex modeling techniques we will cover that will allow us to account for non-stationarity, seasonality effects, and the impact of exogenous variables. Afterwards, we’ll switch from statistical learning techniques to deep learning methods, in order to forecast very large time series with a high dimensionality, a sce- nario in which statist

ML Blog - Everything you need to know about Time Series Forecasting Time series exist in a variety of fields from meteorology to finance, econometrics, and marketing. By recording data and analyzing it, we can study time series to analyze industrial processes or track business metrics, such as sales or engagement. Also, with large amounts of data available, data scientists can apply their expertise to techniques for time series forecasting. We will first learn how to make simple forecasts that will serve as benchmarks for more complex models. Then we will use two statistical learning technique

Explore this link on the map →

related reading