Tiêu đề Day-31-Optimization-Basics-Gradient-Descent ✅

Artificial Neural Network -1
Agenda • Optimization Basics • Gradient Descent Method • Artificial Neural Network (ANN)
Optimization US Federal Aviation Administration(FAA) Challenge: • Responsible for air traffic management • Large-scale weather systems reduce available air space • Developed Airspace Flow Programs to optimize ground delays for each individual flight when weather compromises flight routes Benefits: • Saved airlines $190 million in first 2-years of use Netherland Railways Challenge: • Developed system to optimize 5,500 daily train routes • Creates efficient crew & rolling stock schedules Benefits: • Increased profit by 40 million Euros • Improved on-time arrivals
Preliminaries • Minimize Travel Time, Maximize code coverage, Minimize downtime, Maximize user satisfaction • Profit = Revenue – Expenses • Profit = f (Revenue, Expenses) • Y = f(X1, X2); f(X1, X2) = X1 - X2 • Optimum value of a function f(x) • Related to change, rate of change -> derivatives