Download as pdf or txt
Download as pdf or txt
You are on page 1of 10

Forecasting

CHAPTER 2

NHÓM 2
A qualitative forecasting model

1. New Products or Services

2. Rapidly Changing Markets

3. Long-Term Strategic Planning

4. When Detailed Information is Unavailable

5. Scenario Analysis
The two general
forecasting approaches
Description: Relies on expert judgment and opinions for forecasting,
often used when historical data is scarce or unreliable.
Methods: Market research, Delphi method, panel consensus, scenario
Qualitative Forecasting:
building, market surveys.
Use Cases: New products, emerging markets, uncertain situations.

Description: Relies on historical data and mathematical models,


assuming past patterns continue into the future.
Methods: Time series analysis, regression analysis, econometric
Quantitative Forecasting models, machine learning algorithms.
Use Cases: Industries with reliable historical data, such as finance,
supply chain management, sales forecasting.
Which type of forecast do
you think is the most
appropriate: qualitative or
quantitative? Why?

Qualitative forecast would likely


be more appropriate
Reasoning: Earthquakes are unpredictable events with varying
impacts on communities.
Expert Insight: Qualitative methods allow for expert judgment,
on-the-ground assessments, and inputs from local authorities
and humanitarian organizations.
Adaptability: Qualitative approaches are flexible and can
quickly adjust to changing circumstances and emerging needs
in the aftermath of a disaster.
Is collaborative planning, forecasting, and
replenish­ment (CPFR) applicable in this case?

=> In the immediate aftermath of an earthquake,


Collaborative Planning, Forecasting, and
Replenishment (CPFR) may not apply directly to
forecasting relief aid needs.

However, it could become valuable in later


phases:
Coordination with Relief Organizations.
Logistics and Distribution
Rebuilding Efforts
Long-Term Planning
Moving Averages Weighted Moving Averages Exponential Smoothing

Issue with Trends: Exponential


Issue with Trends: Moving Issue with Trends: While
smoothing places more
averages give equal weight to all weighted moving averages assign
weight on recent data points,
data points within the window, varying weights to different data
but it still considers all past
which can mask underlying points, they still emphasize
data to some extent.
trends. historical values over recent ones.
Effect: This method reacts
Effect: As the average of past Effect: If the trend changes, the
more quickly to recent
data is taken, it lags behind the weighted moving average may
changes compared to simple
actual trend, making it less not quickly reflect this change
moving averages but can still
responsive to changes in the due to the influence of historical
lag behind abrupt or sustained
trend direction. data with higher weights.
trend shifts.
The basic difference between a weighted
moving aver­age and exponential smoothing
A weighted moving aver­age Exponential smoothing

Gives fixed weights to historical data. Each data point has a Adapts the weight assigned to historical data based on the
predetermined weight that does not change over time. importance of the most recent observation. Older data points
have exponentially decreasing weights.
Weighted Moving Average:
Exponential Smoothing

You might also like