Application of Forecasting, Regression and Time Series Models

Overview

ABOUT THE COURSE:Ideas of Forecasting, Regression and Time Series Models help one make accurate forecasts that enabling the decision maker come up with rational as well mathematical and grounded decisions to solve a variety of forecasting, regression and time series problems. This NPTEL course will benefit students in their masters and doctoral programs and working in a variety of areas to tackle and solve interesting problems both from theoretical as well as practical view pointsINTENDED AUDIENCE: 4th Year UG, MTech, MBA, PhDPREREQUISITES: BE/BSc Engg/BTech/BSc/BS-MSINDUSTRY SUPPORT: All consumable, service industry

Syllabus

Week 1:
Introduction of ForecastingSimple Averaging Techniques (fixed/adjusted parameters)Moving Averaging Techniques (fixed/adjusted parameters)Exponential Averaging Techniques (fixed/adjusted parameters)
Week 2:
Exponential Averaging Techniques (fixed/adjusted parameters)Holt Method (single/multiple trend)Holt-Winter Method additive model (single/multiple trend,single/multiple seasonality)
Week 3:
Holt-Winter Method additive model (single/multiple trend, single/multiple seasonality)Holt-Winter Method multiplicative model (single/multiple trend, single/multiple seasonality)Multiple Liner Regression
Week 4:
Simple Liner RegressionMultiple Liner Regression
Week 5:
Multiple Liner Regression (Continuation)Logistic RegressionProbit Model
Week 6:
Ridge RegressionQuantile RegressionLasso Regression
Week 7:
Lasso Regression (continuation)Non-parametric Regression
Week 8:
Polynomial RegressionBayesian Regression (Polynomial Trend, Seasonal, Auto-regression, multi-process, Non-linear Dynamic, Multivariate)
Week 9:
ARMA Models
Week 10:
ARIMA Models
Week 11:
VAR Models
Week 12:
Non-linear time series