دانلود کتاب پیش‌بینی سری‌های زمانی مدرن با پایتون – بررسی مدل‌های آماری، یادگیری ماشین و یادگیری عمیق

Time series data is everywhere: from demand and supply signals, sensor data, financial and operations metrics, to climate, healthcare, and almost any system that varies over time. The challenge is that forecasting involves much more than the selection of a model. It requires understanding patterns, preparing the data appropriately, choosing the right evaluation strategy, and building robust solutions that can be reliably deployed in real-world environments. Modern Time Series Forecasting with Python is written to help you build that full end-to-end capability. The book takes a hands-on approach: we first start with the fundamentals of time series analysis, then gradually build up to classical statistical models, feature-based machine learning, deep learning architectures, and ultimately deployment and forecasting at scale. This book is intended for data scientists, machine learning engineers, analysts, and cloud developers who want to learn forecasting using Python. You should be comfortable with Python fundamentals, and have a basic understanding of statistics. By the end of the book, you will be able to: – Understand time series structure (trend, seasonality, noise) and diagnose stationarity. – Build classical forecasting models (AR, ARMA, ARIMA, SARIMA, exponential smoothing). – Engineer features and train machine learning models for forecasting. – Build deep learning models for sequence data, including LSTMs and more advanced architectures. – Evaluate models correctly using backtesting and time-series-aware validation.