Modern Time Series Forecasting with Python : Industry-ready machine learning and deep learning time series analysis with PyTorch and pandas 2nd edition Manu Joseph by Manu Joseph, Jeffrey Tackes instant download after payment.

Forecasting as a discipline has evolved significantly. For decades, the field was dominated by simple
models that often outperformed more complex ones. Machine learning methods, in various compe-
titions, were repeatedly shown to be uncompetitive or, at best, to add little value. This period, during
which I began my work in forecasting as a PhD student, has been termed the forecasting winter by some.
Since then, much has changed, and we now live in a different world in forecasting. With developments
like the global modeling paradigm and the availability of more data and data with higher frequencies,
machine learning methods have become highly competitive in many forecasting situations, and fore-
casting research is now driven by these approaches. Similarly, on the practitioner side, forecasting
is often carried out by data scientists with a machine learning background but limited specialized
training in forecasting. Their preferred programming tool is usually Python