Mathnal Academy — Supply Chain Analytics Courses
Back to Library
Forecasting With Python Version 2
Forecasting
Bestseller

Forecasting With Python Version 2

by Krish Naidu

English Version 2
₹1,500+ 18% GST
Secure online reader — no downloads. Access anytime from your dashboard.

About This Book

PART I — Classical Statistical Methods 1 Time Series Decomposition: Classical, STL & X-13 2 Exponential Smoothing: SES, Holt's & Holt-Winters 3 ETS State Space: Full Taxonomy of 30 Models 4 Optimising Exponential Smoothing: Grid Search & Log-Likelihood PART II — Demand & ARIMA Models 5 Croston, SBA & TSB: Intermittent Demand Forecasting 6 ARIMA & SARIMA: Box-Jenkins Methodology PART III — Volatility & Financial Models 7 ARCH, GARCH & Volatility Family PART IV — Machine Learning Methods 8 Decision Trees & Random Forests for Time Series 9 Support Vector Regression for Time Series PART V — Bayesian & Probabilistic Methods 10 Bayesian Forecasting: Prior to Posterior 11 Markov Chains & Hidden Markov Models PART VI — Deep Learning Architectures 12 LSTM, GRU & Sequence-to-Sequence Models 13 TBATS, Kalman Filter & Advanced State Space 14 Gaussian Processes for Time Series 15 N-BEATS, N-HiTS & Neural Basis Expansion 16 Temporal Fusion Transformer (TFT) 17 PatchTST, iTransformer & Mamba SSM PART VII — Foundation Models & Production 18 Chronos, TimeGPT & Moirai: Foundation Models 19 DeepAR, Normalising Flows & Density Forecasting 20 Causal Forecasting & Intervention Analysis 21 Production Pipelines, Online Learning & Research Projects

What You'll Learn

Upon successful completion of this book, you will be able to:
Decompose complex time series using Classical Decomposition, STL, and X-13 methods to identify trend, seasonality, cyclical patterns, and irregular components.
Develop and optimize exponential smoothing models, including Simple Exponential Smoothing (SES), Holt's Trend, Holt-Winters, and the complete ETS State Space family of models.
Select and optimize forecasting models using statistical model selection techniques, likelihood estimation, and hyperparameter optimization through Grid Search.
Forecast intermittent and sparse demand using Croston's Method, SBA, and TSB models for inventory management and spare parts forecasting.
Build robust ARIMA and SARIMA models by applying the complete Box-Jenkins methodology, including model identification, parameter estimation, diagnostics, and forecasting.
Model financial volatility and heteroscedasticity using ARCH, GARCH, and their extensions to analyze risk and time-varying variance.
Transform forecasting problems into supervised learning tasks and implement Decision Trees, Random Forests, and Support Vector Regression (SVR) for time series prediction.
Develop Bayesian forecasting models by incorporating prior knowledge, posterior inference, and uncertainty quantification into forecasting workflows.
Model sequential systems using probabilistic state models, including Markov Chains and Hidden Markov Models for regime detection and state transition analysis.
Design and implement deep learning forecasting architectures using LSTM, GRU, and Sequence-to-Sequence (Seq2Seq) networks for complex sequential prediction tasks.
Apply advanced state-space forecasting techniques, including TBATS, Kalman Filters, and modern State Space Models for multiple seasonalities and dynamic systems.
Develop non-parametric forecasting models using Gaussian Processes to estimate uncertainty and model complex nonlinear relationships.
Implement state-of-the-art neural forecasting architectures, including N-BEATS, N-HiTS, Temporal Fusion Transformers (TFT), PatchTST, iTransformer, and Mamba State Space Models.
Leverage foundation models for time series forecasting, including Chronos, TimeGPT, and Moirai, to accelerate forecasting with pre-trained AI models.
Build probabilistic forecasting systems using DeepAR, Normalizing Flows, and density forecasting methods to generate calibrated prediction distributions.
Apply causal forecasting techniques to evaluate interventions, policy changes, promotions, and external events using causal inference methodologies.
Design production-ready forecasting pipelines that support online learning, model monitoring, continuous retraining, and enterprise-scale deployment.
Evaluate and compare forecasting methodologies across statistical, machine learning, Bayesian, deep learning, and foundation model approaches based on accuracy, interpretability, computational efficiency, and business requirements.
Develop end-to-end forecasting solutions for retail, supply chain, finance, manufacturing, energy, healthcare, and other real-world industrial applications.
Apply research-grade forecasting techniques by integrating advanced statistical theory, artificial intelligence, probabilistic modeling, and production engineering into enterprise forecasting systems.
Upon Completion, You Will Be Able To
Master the complete spectrum of modern forecasting methods—from classical statistical models to foundation models for time series.
Build highly accurate forecasting systems using statistical, machine learning, Bayesian, and deep learning techniques.
Develop probabilistic forecasting models that quantify uncertainty and support risk-aware decision-making.
Design production-ready forecasting pipelines capable of continuous learning, monitoring, and deployment in enterprise environments.
Implement state-of-the-art transformer architectures and foundation models for large-scale forecasting applications.
Solve advanced forecasting challenges involving intermittent demand, volatility modeling, hierarchical forecasting, multiple seasonalities, and causal intervention analysis.
Evaluate, compare, and select forecasting models using rigorous statistical principles, machine learning best practices, and business performance criteria.
Deliver research-grade, enterprise-scale forecasting solutions that support strategic planning, operational optimization, and AI-driven decision intelligence.

Reader Reviews

No reviews yet

Be the first to share your thoughts!

Related Books

Forecasting with Python Verson-1
Bestseller
Forecasting

Forecasting with Python Verson-1

by Krish Naidu

₹1,000