Sonal Srivastava is a doctoral student in Management Studies at the Cambridge Judge Business School. Her substantive research interests lie around product development and consumer decision-making in online food and healthcare markets. Methodologically she is interested in combining Behavioural Models, Econometric methods, and Machine Learning to understand decision-making in the presence of large choice sets. As her work spans multiple fields, from economic theory and marketing to algorithm development, she would like to engage and collaborate with different domain experts at Turing to explore new methodologies for her research.
Modelling consumer behaviour is crucial for developing marketing strategy, in designing new products, in assessing customer satisfaction, and in optimising supply-side variables by accurately predicting demand. Theory-based econometric methods, under the discrete-choice framework, are often very useful in modelling consumer decision-making but require extensive, multi-dimensional data, which is seldom available. The advanced Machine Learning (ML) models, on the other hand, are extremely adept at identifying hidden patterns in data but produce what is known as “black-box” predictors which have limited human interpretability. In her Turing research, Sonal will be exploring what is known as the “hybrid” approach in which both- theory-based econometric methods and ML algorithms, are combined to get the best of the two worlds.