How does mathematical optimization work in concert with machine learning?
While machine learning (ML) can enable you to make predictions, mathematical optimization (MO) empowers you to make decisions. When your problem involves complex tradeoffs among various (and often conflicting) business objectives and has an astronomical number of possible solutions, only MO has the power to find the best or optimal solution – which can be used to make optimal business decisions.
Actually, MO and ML are complementary technologies, and more and more companies are developing and deploying applications that combine MO and ML. MO technologies can leverage ML-generated predictions by using them as input for MO-based solutions and decisions. A good example of this is predictive maintenance: ML can enable manufacturers to predict when machine failures are likely to occur, and then MO can use these ML-based predictions to help create optimal maintenance schedules that minimize resource costs and production disruptions. Also, MO-based solutions can be utilized to help shape, retrain, and improve ML models – which can decay over time.
In this webinar, you will hear the results of the 2019 Mathematical Optimization Survey commissioned by Gurobi and conducted by Forrester and gain insights on how data scientists can use tools such as MO in tandem with ML technologies to drive optimal decision-making and business outcomes.