Lecture 4 - Time Series Forecasting Flashcards

1
Q

Define time series forecasting

A
  • technique that tries to predict how a sequence of historical data will continue in the future, by analysing the data and identifying patterns/trends in the series
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

What ways can you identify characteristics in time series data

A
  • visualisation
  • identify patterns, possible explanations for variation in the data
  • statistical analysis
  • for time series –> time plot
  • for seasonal time series –> seasonal plot

–> both can reveal trends, or seasonal behaviour

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Define a time plot

*how can this help in selecting an appropriate time series forecasting approach?

A

depicts the overall trend of the data points across the entire time period.

  • identify trends
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Define a seasonal plot

*how can this help in selecting an appropriate time series forecasting approach?

A

revealing recurring patterns within the data that occur over specific time intervals (seasons)

*recognise seasonality patterns

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Steps of time series forecasting

A
  • Determine time horizon
  • Gather and analyse data

*Select and validate forecasting model to use

  • Make forecast
  • Monitor and control forecast
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

What is a stationary series

A
  • a data sequence which has no strong trend or seasonal component
  • we assume it is essentially constant over the long term with short term fluctuations
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

What forecasting approaches are used for stationary series

A
  • Simple Moving Average
  • Exponential Smoothing
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Key points about simple moving average

A
  • Only user input choice is the length of the moving average

–> short average provides more response to changing demand levels, but may not be desirable

–> long average provides more smoothing, but may miss trends and turning points

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

For a 3 week simple moving average, where does the first moving average point go

A

In the ‘3 week moving average’ column, in line with the 4th week’s data

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Define bias in time series forecasting

A

a forecast is biased if they consistently overestimate or underestimate values

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

Why is average error not a good measurement of forecast accuracy (error and bias)

A

Because negative and positive errors will cancel out

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Errors made consistently in one direction imply ….. what?

A

bias

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Three common measures of TIME SERIES forecast accuracy

A

*Mean Absolute Deviation (MAD)

*Mean Squared Deviation/Error (MSD / MSE)

*Mean Absolute Percentage Error (MAPE)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

What is MAD

A

*Mean Absolute Deviation

*measures the average of the absolute differences between forecasted values and actual values

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

How is ‘Mean Absolute Deviation’ calculated

A

𝗠𝗔𝗗 = Σ | 𝗙𝗼𝗿𝗲𝗰𝗮𝘀𝘁 - 𝗔𝗰𝘁𝘂𝗮𝗹 | / 𝗻

with Σ (sigma) representing the sum over all n data points.

&

|…| representing absolute value

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

Strengths and weaknesses of MAD as a metric to measure error and bias

A

Strength
* easy to interpret as in sameΣ | (Forecast - Actual) / Actual | * 100% / n scale with the data
* less sensitive to outliers than MSE

Weakness
*doesn’t consider magnitude of errors (as mean treats all errors equally)

17
Q

What is MSD/MSE

A

*Mean Squared Deviation/Error

*measures the average of the squared differences between the forecasted values and the actual values

  • squaring the differences gives more weight to larger errors
18
Q

How is ‘Mean Squared Deviation/Error’ calculated

A

Σ (Forecast - Actual)² / n

19
Q

Strengths and weaknesses of MSD/MSE

A

Strengths
*more sensitive to larger errors (due to squaring)

Weaknesses
* results can be skewed by outliers
*can be difficult to interpret error without converting back to original scale of data

20
Q

What is MAPE

A
  • Mean Absolute Percentage Error
  • measures the average of the absolute percentage errors
21
Q

How is ‘Mean Absolute Percentage Error’ calculated

A

Σ | (Forecast - Actual) / Actual | x 100% / n

22
Q

Strengths and Weaknesses of MAPE

A

Strengths
*easily interpreted as a %
–> therefore useful for comparison across data sets 𝘄𝗶𝘁𝗵 𝗱𝗶𝗳𝗳𝗲𝗿𝗲𝗻𝘁 𝘀𝗰𝗮𝗹𝗲𝘀.

Weaknesses
* can be misleading when demand levels are very low (close to 0 would mean dividing by 0’s)
* sensitive to outliers

23
Q

What is exponential smoothing

A

forecasting technique that assigns exponentially decreasing weights to past observations when generating forecasts

  • assigns higher weights to more recent data points and lower weights to older ones, therefore giving them less influence on the forecast.
24
Q

A higher α (closer to 1) …..?

A

*less weight on past observations

*more responsive curve to changes in data

25
A lower α (closer to 0) .....?
* more weight on historical data * more smooth curve, less sensitive to changes in data
26
Strengths and weaknesses of exponential smoothing
Strengths * practical forecasting method --> easy to use * focus on recent data and current situation Weaknesses * difficulty handling complex data with multiple trends/relationships * subjective selection of smoothing factor (may require experimentation to find best parameter --> time consuming) * lags behind trends, if one is present
27
for practical reasons, what's the best smoothing factor to use
(0.3 - 0.1) because provides a stable series
28
Why§ Why is it not a good idea to use (simple) exponential smoothing or moving averages for a time series with a significant trend effect?
* struggle to capture trends as the importance of older data points decreases over time --> these points may hold important information regarding the trend * both gradually converge to a constant value --> predicting that the pattern in the series will eventually flatten out * weighting makes it slow to react to significant changes in trend
29
Approaches to use when time series has a trend and when the time series has a seasonality factor
TREND * Trend extrapolation (linear regression) * Holt's method (double exponential smoothing) SEASONALITY * De-seasonalise * Holt-Winters Technique (triple exponential smoothing)
30
Describe how to perform trend extrapolation (linear regression)
find the line of best fit (need to know intercept and slope) y = mx + c least square method - sum of vertical distances between predicted and actual demand , all squared
31
Describe Holt's Method
(for linear trends) * exponential smoothing with two smoothing factors --> one focus on smoothing the base (intercept) with weighting the most recent observation like in exp. smoothing --> second focus on the trend (slope) based on the difference between most recent forecast and previous level estimate
32
How to de-seasonalise a time series with a seasonality factor
* calculate sample mean of all values * divide each observation by sample mean, to find initial seasonal factors * average the factors for like periods (e.g. mean of all mondays, mean of all tuesdays and so on) - this finds the seasonal factors * FORECAST (e.g. forecast for Tuesdays) - sample mean x seasonal factor for Tuesdays
33
Describe Holt-Winter's technique
(triple exponential smoothing) ‘De-trend’ and ‘de-seasonalise’ the time series by separating base from trend and seasonality effects * base (intercept) smoothing * trend (slope) smoothing * seasonal smoothing (with seasonal co-efficient)
34
Some key further difficulties with time series forecasting
badly behaved series don't allow for accurate forecasts * Intermittent and erratic demand
35
What is forecast automation in time series forecasting
the use of software and tools to automate the process of generating forecasts for future values in a time series
36
Why is forecast automation important for companies that deal with many SKUs (Stock Keeping Units)? any challenges?
𝗜𝗺𝗽𝗼𝗿𝘁𝗮𝗻𝘁 𝗕𝗲𝗰𝗮𝘂𝘀𝗲.. * more efficient and accurate as less manual intervention * scalable * can update easily with new trends or data * consistent across all stock units * data driven insights can inform pricing and marketing strategies 𝗖𝗵𝗮𝗹𝗹𝗲𝗻𝗴𝗲𝘀... * human judgement required to adjust forecasts because of internal (price changes, promotions) or external factors (economy, actions of competition) --> also outlier spotting is done by humans