Supply chain optimization

In order to manage a supply chain, it is necessary to be able to optimally meet the projected demand with enough buffer to minimize out-of-stock events without holding excessive inventory. Using demand forecasts and information about the supply chain and products will allow for the supply chain to be managed to maximum profit.

Overview

Find optimal levels of product throughout the supply chain to maximize profit and minimize out-of-stock events.

TARGET

No target, this is an optimization of profit under multiple conditions.

Challenge

Starting with an existing demand forecast, information on shipments, times, warehouses, likely weather and external events will need to be aggregated into a final optimization problem. This data is likely to sit within multiple databases owned by different organizations within the company.  In addition, external data on weather and events will help improve this optimization. All this data needs to be cleaned, joined and transformed into valuable ML features before going into model training. This pre-modeling prep process can be frustrating and time consuming. We are here to help.

Modeling techniques and libraries

Supply chain optimization

Build and solve the constrained optimization problem to find the optimal inventory levels throughout the supply chain that maximizes profit while minimizing out of stock issues.

Package:
  • CVOXPT
  • pyOpt
  • scipy.optimize

Data features

Customer Project cost
ERP
Data Type
Continuous
Target
No
Yes
Customer Shipments
ERP
Data Type
Continuous
Target
No
Yes
Day of Week
Calendar
Data Type
Categorical
Target
No
Yes
Event
Calendar
Data Type
Binary
Target
No
Yes
Fulfillment cost
Inventory
Data Type
Continuous
Target
No
Yes
Headcount assigned
ERP
Data Type
Continuous
Target
No
Yes
Holiday
Calendar
Data Type
Binary
Target
No
Yes
Inventory cost
Inventory
Data Type
Continuous
Target
No
Yes
Number of Stores carrying product
Inventory
Data Type
Continuous
Target
No
Yes
Number of Stores carrying related product
Inventory
Data Type
Continuous
Target
No
Yes
Number of Stores out of stock
Inventory
Data Type
Continuous
Target
No
Yes
Online product cost
ERP
Data Type
Continuous
Target
No
Yes

Related accelerators

No items found.

Get your data science on.