Methods and apparatus to reduce signal-to-noise ratio (snr) of monadic scores

Patent metadata

Field	Value
Publication number	US-2025348895-A1
Application number	US-202519219759-A
Country	US
Kind code	A1
Filing date	May 27, 2025
Priority date	Mar 12, 2021
Publication date	Nov 13, 2025
Grant date	—

Abstract

Methods and apparatus disclosed herein reduce signal-to-noise ratio (SNR) of monadic scores. An example apparatus to reduce a signal-to-noise ratio (SNR) of monadic scores, the apparatus includes memory, machine readable instructions, and processor circuitry to execute the machine readable instructions to at least identify a discrete choice probability of selection corresponding to a first product, generate a scale question corresponding to the first product, calculate a monadic probability corresponding to the first product based on the scale question for the first product, and reduce the SNR of the monadic probability by joining the discrete choice probability of selection of the first product with the monadic probability of selecting the first product.

First claim

1 .- 20 . (canceled)

21 . An apparatus, the apparatus comprising:
interface circuitry; machine readable instructions; and at least one processor circuit to be programmed by the machine readable instructions to:
generate a graphical user interface (GUI) to present market-available products on a display device;
generate choice sets of the market-available products, the choice sets associated with a candidate product;
retrieve, from the GUI, selection information corresponding to (a) the market-available products and (b) the candidate product;
calculate a discrete choice probability of selection based on the selection information corresponding to a first one of the market-available products by:
applying a Gumbel distribution to a deviation between (1) an observed outcome and (2) a true outcome;
associating one or more random variables, with an ordered logit model, to a binomial random variable; and
modifying a data structure of a memory with distributed ones of the one or more random variables based on the Gumbel distribution, the one or more random variables distributed in the data structure based on (a) a probability density function and (b) a cumulative distribution function;

calculate a monadic probability corresponding to the first one of the market-available products based on the Gumbel distribution, the discrete choice probability of selection and the monadic probability associated with a homogenous population of panelists;
generate a combined likelihood of selection by joining the discrete choice probability of selection of the first one of the market-available products with the monadic probability of selecting the first one of the market-available products;
generate a signal-to-noise ratio (SNR) based on data inconsistencies corresponding to a likelihood of product selection associated with the combined likelihood of product selection; and
cause one or more products to be released for public access by adjusting to a customer probability of purchasing the first one of the products.

22 . The apparatus of claim 21 , wherein one or more of the at least one processor circuit is to calculate the monadic probability corresponding to the first one of the products based on the Gumbel distribution and a cut-off point error.

23 . The apparatus of claim 22 , wherein the cut-off point error is associated with an average item utility value distribution.

24 . The apparatus of claim 21 , wherein one or more of the at least one processor circuit is to generate a combined likelihood of product selection based on at least one of a distribution function of utility or a probability of item selection.

25 . The apparatus of claim 24 , wherein one or more of the at least one processor circuit is to automatically select at least one of the distribution function of utility or the probability of item selection based on at least one of a K-point Likert scale or a discrete choice-based selection, respectively.

26 . The apparatus of claim 21 , wherein one or more of the at least one processor circuit is to reduce the SNR of the monadic probability using a scaling factor indicative of a type of correlation between the discrete choice probability and the monadic probability, the correlation being a positive correlation or a negative correlation.

27 . The apparatus of claim 21 , wherein one or more of the at least one processor circuit is to perform a post-estimation analysis of data inconsistencies on an individual level, a group level, or an aggregate level.

28 . The apparatus of claim 21 , wherein the choice sets are associated with a future product or a current in-market product.

29 . The apparatus of claim 21 , wherein the products are released for public access by adjusting a volume of the one or more products a manufacturer sells in a post-product launch.

30 . The apparatus of claim 21 , wherein the observed outcome corresponds to an expected utility of a selected product and the true outcome corresponds to an actual utility of the selected product.

31 . The apparatus of claim 21 , the observed outcome corresponding to an expected utility of a selected product and the true outcome corresponding to an actual utility of the selected product.

32 . A computer implemented method to reduce a signal-to-noise ratio (SNR) of monadic scores, the method comprising:
generating a graphical user interface (GUI) to present market-available products on a display device; generating choice sets of the market-available products, the choice sets associated with a candidate product; retrieving, from the GUI, selection information corresponding to (a) the market-available products and (b) the candidate product; calculating a discrete choice probability of selection based on the selection information corresponding to a first one of the market-available products by:
applying a Gumbel distribution to a deviation between (a) an observed outcome and (b) a true outcome;
associating one or more random variables, with an ordered logit model, to a binomial random variable; and
modifying a data structure of a memory with distributed ones of the one or more random variables based on the Gumbel distribution, the one or more random variables distributed in the data structure based on (a) a probability density function and (b) a cumulative distribution function;
calculating a monadic probability corresponding to the first one of the market-available products based on the Gumbel distribution, the discrete choice probability of selection and the monadic probability associated with a homogenous population of panelists; generating a combined likelihood of selection by joining the discrete choice probability of selection of the first one of the market-available products with the monadic probability of selecting the first one of the market-available products; generating a signal-to-noise ratio based on data inconsistencies corresponding to a likelihood of product selection associated with the combined likelihood of product selection; and causing one or more products to be released for public access by adjusting to a customer probability of purchasing the first one of the products.

33 . The computer implemented method of claim 32 , further including calculating the monadic probability corresponding to the first one of the products based on the Gumbel distribution and a cut-off point error.

34 . The computer implemented method of claim 33 , wherein the cut-off point error is associated with an average item utility value distribution.

35 . The computer implemented method of claim 32 , further including generating a combined likelihood of product selection based on at least one of a distribution function of utility or a probability of item selection.

36 . The computer implemented method of claim 32 , further including automatically selecting at least one of a distribution function of utility or a probability of item selection based on at least one of a K-point Likert scale or a discrete choice-based selection, respectively.

37 . The computer implemented method of claim 32 , further including reducing the SNR of the monadic probability using a scaling factor indicative of a type of correlation between the discrete choice probability and the monadic probability, the correlation being a positive correlation or a negative correlation.

38 . At least one non-transitory machine-readable medium comprising machine-readable instructions to cause at least one processor circuit to at least:
generate a graphical user interface (GUI) to present market-available product

Assignees

• Nielsen Consumer Llc

Inventors

• Malek Kamal M
• Kebarighotbi Ali
• Teller David B

Classifications

• G06N7/01

  Probabilistic graphical models, e.g. probabilistic networks · CPC title

• G06Q30/0203

  Market surveys; Market polls · CPC title

• G06Q30/0201 Primary

  Market modelling; Market analysis; Collecting market data · CPC title