Complex Predictive Modelling & Forecasting

Day 1
Module 1: Introduction to Advanced Predictive Modelling

Module 2: Revisit Time Series Methods (ACF/PACF Functions, AR/MA)

Module 3: ARIMA & Seasonal ARIMA Methods

Module 4: Workshop 1: Forecasting using ARIMA/SARIMA methods based on relevant practical case study
Day 2
Module 5: Extending Univariate to Multivariate Time Series – Transfer Functions

Module 6: Introduction to ARCH & GARCH Modelling

Module 7: Workshop 2: Time Series Forecasting case study using Transfer Functions

Quiz 1
Day 3
Module 8: Introduction To Conjoint Analysis

Module 9: Traditional Conjoint

Module 10: Adaptive Conjoint Analysis (ACA)

Module 11: Workshop 3: Case study: Traditional Conjoint Models development to Solve an Industry Problem
Day 4
Module 12: Choice-Based Conjoint (CBC)

Quiz 2

Module 13: Predictive modelling Using Survival Analysis

Module 14: Workshop 4: Case Study & Workshop using CBC & ACA to Solve an Industry Problem
Day 5
Module 15: Survival Analysis continued

Module 16: Case Study and Workshop on Survival Analysis Modelling

Quiz 3

Topic 1:
Advanced Time Series Forecasting: There are some complex industry forecasting problems which can’t be solved using time series regression,dummy variable regression, decomposition methods, Variations of Exponential Smoothing & ARIMA. Much more sophisticated methods are needed to solve some of the problems that arise in the industry.

We focus on some of the advanced versions of ARIMA: SARIMA, ARIMAX/ Transfer Function Models. ARCH & GARCH Modelling for finance-related modelling
Topic 2:
Conjoint Analysis: Conjoint analysis is a statistical method for finding out how consumers make trade-offs and choose among competing products or services. It is also used to predict (simulate) consumers’ choices for future products or services.
There are three types of methods available: 1) Traditional Conjoint 2) Adaptive Conjoint and 3) Choice-Based Conjoint. This advanced technique has been used in market research for a long time. It is an effective tool to improve service quality, enhance product features, understand competition and predict market share.
Topic 3:
Survival Analysis: In the past, this topic used to come under reliability theory primarily used in the manufacturing sector and biomedical industry. The technique was used to predict the lifetime of machines as well as humans. It is an area of predictive analytics where the dependent variable is truncated and hence requires special treatment. The application of this Methodology is also known as Time to Event Modelling. In the modern world, this Methodology is used to predict many scenarios like playtime prediction of popular games, choosing stocks for investment decision by a firm’s performance in stressed situation etc.

Add to Comparison Back to All Courses