SOLUTIONS |
Supply Chain Management
Demand Forecasting Solution
TAKE CAMEO
Computational Algorithms for Managing Efficiency of Operations
 Controlling the supply chain starts with the right forecast of future demand. Internal processes can be optimized perfectly, but if estimations of the future demand are incorrect you will have inaccurate information.
Demand Forecasting
Demand forecasting is the process of determining what, where, when, and in what quantities products are needed. Accurate demand forecasting can give you a significant competitive advantage.
There are three types of forecasting;.
- Qualitative: In this type, forecasting is done through judgmental, intuition and informed opinions of experts. This type of forecasting is done during the introduction of new products when demand is not yet known.
- Extrinsic: In this type, projections are based on external (extrinsic) indicators. Research shows that the demand for a product group correlates to activity in corresponding field. For example sales of automobile tires are proportional to fuel consumption.
- Intrinsic: In this type, historical data is used to predict future data. Several specific techniques are used including moving averages, and exponential smoothing
Artificial Neural Network (ANN)
A neural network is a massively parallel distributed processor that has a natural propensity for storing experiential knowledge and making it available for use. It resembles the brain in two respects:
- Knowledge is acquired by the network through a learning process.
- Inter neuron connection strengths known as “weights” are used to store knowledge.
ANN’s have a "training" rule whereby the weights of connections are adjusted on the basis of data. In other words, NNs "learn" from examples and exhibit a capability for generalization beyond the training data.
ANNs normally have great potential for parallelism, since the computations of the components are largely independent of each other. Some people regard massive parallelism and high connectivity to be defining characteristics of NNs, but such requirements rule out various simple models, such as simple linear regression which are usefully regarded as special cases of NNs.
How forecasting is done in TAKE CAMEO through ANN?
Manufacturers and retailers can forecast aggregate demand figures with some certainty, but it is becoming increasingly difficult to predict how the demand will be distributed among the many SKUs they sell. Sales forecasting is further complicated owing to influence of internal and external environments.
Artificial neural networks (ANNs) have come up as a powerful technique to learn time series data since their performances in the area of control and pattern recognition.
TAKE Cameo’s Forecasting engine uses artificial neural network by sigmoid function employing a feed-forward ANN with the error back-propagation (EBP) learning algorithm. The values of monthly sales and special actions like promotion schemes are used as inputs to the system. The lagged values of sales of at least a business cycle (or financial year) should be available.
To receive more information on TAKE CAMEO click here.
Demand Forecasting Solution Data Sheet Download
< Back to Supply Chain Planning Solutions
|
