Goal

Develop an algorithm for calculating the optimal volume of goods delivery to the warehouse. Decrease the cost of the goods in the warehouse without reducing the current sales volume.

Our solution

Data on goods delivery and sales for the past year were used as the initial data.

Probability distributions of goods discounts emergence were created on the basis of the initial data.

Numerous samples were created based on the general trends identified from the original data. Warehouse cost at different times became the elements of those samples. Thus, the random variables influence was minimized.

Each of the samples was visualized. The blue line indicates the actual data, the orange line indicates the calculations results.

For each sample, optimization of the goods delivery to the warehouse was carried out. After that, the values ​​obtained were averaged.

The final chart representing changing of the warehouse cost.

The final chart representing the warehouse cost.

The blue line indicates the actual data, the orange line indicates the calculations results based on the initial state of the warehouse at the beginning of the period, and the green line indicates the calculations results based on the optimal state of the warehouse.