Lecture 1
General Concepts
1. What is the primary goal of data analytics?
2. Define and di:erentiate between data science, data analytics, and data mining.
3. What are the "3Vs" associated with big data?
Data Analytics Process
4. What are the six key steps of the data analytics process?
5. Why is data pre-processing important in the data analytics pipeline?
6. What is the purpose of data exploration, and which tools can be used for this
step?
Analytics Types
7. Compare and contrast descriptive analytics and diagnostic analytics.
8. What is the main focus of predictive analytics, and how does it di:er from
prescriptive analytics?
Data Types
9. What is the di:erence between cross-sectional data and time series data?
Provide an example for each.
10. How are time series components such as trend and seasonality identified and
defined?
Forecasting
11. What is the purpose of forecasting in the context of data analytics?
12. Explain the di:erence between one-step-ahead and multi-step-ahead forecasts.
Pre-processing Techniques
13. What are some common issues addressed during data cleaning?
14. Describe the di:erence between normalization and standardization during data
transformation.
Practical Applications
15. How can data analytics benefit supply chain management in terms of logistics
and demand forecasting?
,Lecture 2
Partitioning Series
1. What is the purpose of partitioning a dataset into training and validation sets?
2. Describe the consequences of not partitioning a dataset.
3. Define overfitting and underfitting. How do these a:ect a forecasting model?
Methods for Partitioning Data
4. Explain the concept of a simple train-test split and its typical data ratio.
5. How does rolling window cross-validation di:er from expanding window cross-
validation?
6. What factors influence the choice of training and validation periods?
Forecasting Methods
7. What are the key classifications of forecasting methods, and how do they di:er?
8. Di:erentiate between judgmental methods and data-driven methods of
forecasting.
9. Compare univariate methods to multivariate methods in forecasting.
Performance Evaluation
10. Why is it necessary to evaluate and compare forecasting models?
11. List and describe at least three performance metrics used to evaluate
forecasting models.
12. Which metric penalizes larger forecasting errors more heavily: Mean Squared
Error (MSE) or Mean Absolute Error (MAE)? Why?
Smoothing Methods
13. What is the main goal of smoothing in time series analysis?
14. Di:erentiate between moving averages and exponential smoothing techniques.
15. How is a trailing moving average calculated, and when might it be used?
, Lecture 3
General Concepts
1. What is the purpose of smoothing methods in time series analysis?
2. Di:erentiate between centered and trailing moving averages.
3. What is the naive forecast method, and when is it used?
Moving Average
4. How is the trailing moving average calculated, and how does it di:er from the
centered moving average?
5. Why can’t the centered moving average be used for forecasting?
6. What are the advantages and disadvantages of using a smaller versus a larger
window width ww?
Simple Exponential Smoothing
7. What assumptions does simple exponential smoothing make about a time
series?
8. How does the smoothing parameter αα a:ect forecasts in exponential
smoothing when it is close to 0 versus when it is close to 1?
9. Write the formula for simple exponential smoothing in its weighted average and
component forms.
Advanced Smoothing Methods
10. What are the limitations of simple smoothing methods, and how are they
addressed by Holt’s model and Holt-Winter’s model?
11. What are the key di:erences between additive and multiplicative trends in Holt’s
model?
12. How does Holt-Winter’s model incorporate seasonality into its forecasts?
Practical Applications
13. What is the ETS framework in smoothing methods, and what do its components
represent?
14. How are the initial values for l0l0 (level) and b0b0 (trend) estimated in Holt’s
model?
15. Using Holt-Winter’s model, how do you forecast a time series with both trend
and seasonality?
General Concepts
1. What is the primary goal of data analytics?
2. Define and di:erentiate between data science, data analytics, and data mining.
3. What are the "3Vs" associated with big data?
Data Analytics Process
4. What are the six key steps of the data analytics process?
5. Why is data pre-processing important in the data analytics pipeline?
6. What is the purpose of data exploration, and which tools can be used for this
step?
Analytics Types
7. Compare and contrast descriptive analytics and diagnostic analytics.
8. What is the main focus of predictive analytics, and how does it di:er from
prescriptive analytics?
Data Types
9. What is the di:erence between cross-sectional data and time series data?
Provide an example for each.
10. How are time series components such as trend and seasonality identified and
defined?
Forecasting
11. What is the purpose of forecasting in the context of data analytics?
12. Explain the di:erence between one-step-ahead and multi-step-ahead forecasts.
Pre-processing Techniques
13. What are some common issues addressed during data cleaning?
14. Describe the di:erence between normalization and standardization during data
transformation.
Practical Applications
15. How can data analytics benefit supply chain management in terms of logistics
and demand forecasting?
,Lecture 2
Partitioning Series
1. What is the purpose of partitioning a dataset into training and validation sets?
2. Describe the consequences of not partitioning a dataset.
3. Define overfitting and underfitting. How do these a:ect a forecasting model?
Methods for Partitioning Data
4. Explain the concept of a simple train-test split and its typical data ratio.
5. How does rolling window cross-validation di:er from expanding window cross-
validation?
6. What factors influence the choice of training and validation periods?
Forecasting Methods
7. What are the key classifications of forecasting methods, and how do they di:er?
8. Di:erentiate between judgmental methods and data-driven methods of
forecasting.
9. Compare univariate methods to multivariate methods in forecasting.
Performance Evaluation
10. Why is it necessary to evaluate and compare forecasting models?
11. List and describe at least three performance metrics used to evaluate
forecasting models.
12. Which metric penalizes larger forecasting errors more heavily: Mean Squared
Error (MSE) or Mean Absolute Error (MAE)? Why?
Smoothing Methods
13. What is the main goal of smoothing in time series analysis?
14. Di:erentiate between moving averages and exponential smoothing techniques.
15. How is a trailing moving average calculated, and when might it be used?
, Lecture 3
General Concepts
1. What is the purpose of smoothing methods in time series analysis?
2. Di:erentiate between centered and trailing moving averages.
3. What is the naive forecast method, and when is it used?
Moving Average
4. How is the trailing moving average calculated, and how does it di:er from the
centered moving average?
5. Why can’t the centered moving average be used for forecasting?
6. What are the advantages and disadvantages of using a smaller versus a larger
window width ww?
Simple Exponential Smoothing
7. What assumptions does simple exponential smoothing make about a time
series?
8. How does the smoothing parameter αα a:ect forecasts in exponential
smoothing when it is close to 0 versus when it is close to 1?
9. Write the formula for simple exponential smoothing in its weighted average and
component forms.
Advanced Smoothing Methods
10. What are the limitations of simple smoothing methods, and how are they
addressed by Holt’s model and Holt-Winter’s model?
11. What are the key di:erences between additive and multiplicative trends in Holt’s
model?
12. How does Holt-Winter’s model incorporate seasonality into its forecasts?
Practical Applications
13. What is the ETS framework in smoothing methods, and what do its components
represent?
14. How are the initial values for l0l0 (level) and b0b0 (trend) estimated in Holt’s
model?
15. Using Holt-Winter’s model, how do you forecast a time series with both trend
and seasonality?
