L7: Time Series Forecating

Question 1

What is the goal of time series forecating?

Answer 1

Predict future state of values based on historical values

Question 2

What is unique about time series forecasting vs. regression or classification?
A) Time dependency and sequential data rather than individual draws
B) Feature engineering: use lag values (past values) as features and time-based features
C) Different evaluation metrics: mean absolute error, root mean squared error etc.
D) Time-based cross validation or walk-forward validation
E) Interdependence between observations
F) Different data partitioning: test and training “periods” rather than “data”
G) Captures seasonality and trends

Answer 2

All options are true

Question 3

The longer the horizon, the larger the uncertainty of the forecast, leading to the benefit of forecast updating

TRUE/ FALSE

Answer 3

TRUE

Question 4

Time series ANALYSIS is descriptive as it describes the patterns in historical data.

Time series FORECASTING is predictive, as it predicts future states of values

Question 5

What does residual error mean in times series forecasting?

Answer 5

Residual error: difference between forecast value vs. actual value

Question 6

What does “TREND” in times series mean?

Answer 6

Whether the series has an overall increasing/ decreasing trend

Question 7

What does “SEASONALITY” in times series mean?

Answer 7

Whether any spikes repat itself at fixed intervals (seasonality) or non-fixed interval (cycle)

Question 8

What does “RANDOM NOISE” in times series mean?

Answer 8

Any form of observation left after accounting for trend and seasonality

Question 9

What is global trend vs. local trends?

Answer 9

Local trend: Represents shorter-term fluctuations or patterns within data

Global trend: represents the overall long-term direction of data - captures the broad movement/ pattern

Question 10

Exponential Smoothing means that observations in the past are weighted depending on how far back they are in time (closer in time observations are assigned higher weights).

TRUE/FALSE

Answer 10

TRUE

Question 11

The degree of weighting depends on the domain: whether it is fast-changing or fairly constant over time

TRUE/FALSE

Answer 11

TRUE

The degree of decay of weights back in time depends on how important newer data is

Question 12

Data partitioning in time series forecast: e.g., fixed partitioning implies that you use all past data as training dataset to forecast the future

TRUE/FALSE

Answer 12

TRUE

Question 13

What are the common data pre-processing considerations in TSF? Select all correct
A) Missing values: some models, e.g., ARIMA cannot handle missing values
B) Unequally spaced series: equal intervals between datapoints is necessary, if not, imputation is required (e.g., setting N.A. to 0)
C) Choice of time span: older data is less informative, but important to have sufficient dataset size

Answer 13

ALL ARE CORRECT:

A) Missing values: some models, e.g., ARIMA cannot handle missing values
B) Unequally spaced series: equal intervals between datapoints is necessary, if not, imputation is required (e.g., setting N.A. to 0)
C) Choice of time span: older data is less informative, but important to have sufficient dataset size

Question 14

Sliding window is one partitioning technique in TSF. What does this mean?
A) Older data is dropped due to the assumption that newer data is more representative - smaller training set
B) Longer training periods are created as more data is revealed with to be incorporated into the model

Answer 14

A) Older data is dropped due to the assumption that newer data is more representative - smaller training set

Question 15

Expanding window is one partitioning technique in TSF. What does this mean?
A) Older data is dropped due to the assumption that newer data is more representative - smaller training set
B) Longer training periods are created as more data is revealed with to be incorporated into the model

Answer 15

B) Longer training periods are created as more data is revealed with to be incorporated into the model

Question 16

The automatically generated Naive benchmark in TSF represents what value?

Answer 16

The latest observed value

Question 17

What does sNaive stand for?
It gives us the last seasonal value (i.e. we predict something in December, it will give us the last observation in December last year

Answer 17

Seasonal naive

Question 18

In TSF, you want to extract all information from the data, leaving behind only white noise.
To check if this is accomplished, you can use which measure?
A) autocorrelation function
B) autocorrection function
C) autocausation function

Answer 18

A) autocorrelation function
The higher the ACF, the more data is left on the table-

If most information has been extracted in the model, the ACF is very low, indicating low interdependency