I Conducted 27 experiments for a 10-factor,3-level design of experiments. I conducted these experiments on a decentralized production-distribution supply chain on a desktop computer to understand the effect (importance) of each of the 10 factors on the total supply chain cost (Response).
These experiments are computer simulations.
The 10 factors considered were namely, Tardiness cost (shortage cost at customer zone) ,(TarC), Earliness cost at the Distribution centre (EarC), Transportation cost between factories and Distribution centers (TransFDC), Transportation cost between Distribution centers and Customer Zones (TransDCCU), Production Cost of Products at Factory (ProdC), Inventory holding cost at factory (InvC), Unfulfilled cost (Shortage cost at factory (UnfulfillC), capacity of factory (CapaF), Maximum earliness allowed (Capae), Inventory holding capacity at factory(Capaif)
For each of the 10 factors, I have chosen 3-levels (1,2,3).
1-represents low,2-represents medium, 3-represents high.
I have assigned values for these 3 levels for each of the 10-factors as shown in Table.1(Attachment) . For instance for Tardiness Cost (1stfactot).,i.e, TarC, level 1 represents a value between 100 and 500. For the same TarC , level-2 represents a value between 500 and 1000. Level-3 for TarC represents a value between 1000 and 5000. Like above, the 3-level values have been assigned for all the 10-factors.
The settings of parameters for conducting experiments have been shown in Table.2.(Attachment)
The Total cost (Table.3) of the Decentralized supply chain has been obtained by solving the above 10-factor experiments using a G.A.M.S. modeling language for optimizing the cost. The total cost (Response) for the 27 experiments was shown in Table.3.(Attachment)
I want to implement the above inputs(factors), Targets(Responses) in Artificial Neural Network of MATLAB to know the impact (importance) of each of the 10-factors on total cost (Response) of the decentralized supply chain.
can somebody help me in implementing the data shown in the attachment for using ANN to know the importance of each of the 10 factors on total supply chain cost