YieldWise provides research and development services for corporate clients seeking top experts for cutting-edge analytical projects. Our research group keeps YieldWise’s analytical and software development technologies and services at the forefront of the industry.

Below are just some of our research and development projects:

  • Development of a software solution for advanced survey analytics using a polytomous Rasch measurement model, inter-rater agreement, and Bayesian Relational Networks. This solution can be used for workforce planning, talent management, consumer insights and market research (consumer and product profiles).
  • Development and deployment of a system that monitors borrowers’ financial behavior and creates optimal actions to prevent loan defaults (design of statistical experiments, machine-learning classification using support vector machine method and its probabilistic variations, time-to-event models).
  • Estimation of instantaneous risk and prediction of losses (Bayesian Cox hazard model, asset depreciation based on Gompertz-like model).
  • Design of large-scale statistical experiments to identify active factors that impact customer purchasing behavior (Plackett-Burman designs followed up with fractional factorial plans).
  • Estimation of price, supply, and demand elasticities to optimize prices per store/product (Hidden Markov model, support vector machines, design of statistical experiments).
  • Development of a large-scale product recommendation system using Bayesian Relational Networks.
  • Development of a systematic approach for discovering the source of faults that cause yield degradation in the manufacturing process of semiconductor IC (integrated circuits) devices. 
  • Development of discrimination criterion for finding the active factors in screening experiments.
  • Development of a systematic approach intended to reduce the purchase cost of Web proxy caches by optimal using of available resources.
  • Development and deployment of complex statistically designed digital and TV advertising experiments.
  • Development of models, algorithms, and programs for digital content recommendation and personalization.
  • Development of solution for behavior nano-clustering of customers using Relational Bayesian Networks methodology.
  • Development and deployment of advanced statistical methods and algorithms for measurement of Search and Advertising processes such as web search, mobile search, shopping search, etc.
  • Development and implementation of advanced methods and algorithms for discovering and analyzing the commercial intentions of web visitors.
  • Development of predictive models of customers purchasing behavior (time-to-event predictions).
  • Development of methods and algorithms for predicting the popularity of messages on X (former Twitter). The implementation of the system is intended to indicate whether or not a message will be retweeted and predict the volume of retweets/reposts.