Assortment
2 min read
Updated:
- Microsoft Azure@SKU optimization for consumer brands
- Parametric model:
- They used multinomial logistic regression to model the probability that a consumer chooses an item at a specific time, given a set of items of that cateogry in an assortment with a known utility to the customer.
- They also assume that the utility of an item can be a function of its features. External information can also be included in the measure of utility (for example, an umbrella is more useful when it rains).
- Neural networks with a softmax output layer could be used effectively on large multi-class problems to replace the multinomial logistic regression.
- Nonparametric model:
- MNL assumes that the relative probability of someone choosing between two options is independent of additional alternatives introduced in the set later. That’s impractical in most cases.
- For instance, if you like product A and product B equally, you’ll choose one over the other 50% of the time. Let’s introduce product C to the mix. You may still choose product A 50% of the time, but now you split your preference 25% to product B and 25% to product C. The relative probability has changed.
- Also, MNL and derivatives have no simple way to account for substitutions that are due to stock-out or assortment variety (that is, when you have no clear idea and pick a random item among those on the shelf).
- Non-parametric models are designed to account for substitutions and impose fewer constraints on customer behavior.
- They introduce the concept of ranking, where consumers express a strict preference for products in an assortment. Their purchasing behavior can therefore be modeled by sorting the products in descending order of preference.
- The assortment optimization problem can be expressed as maximization of revenue
- In such a formulation, the problem can be regarded as a mixed-integer optimization.
- Parametric model: