Florence Chi
St. Paul’s Academy
Abstract
Optimization has played a significant role in training neural
networks [23]. This has resulted in a number of efficient algorithms
[22, 3, 5, 29, 31] and practical applications in medical diagnosis
and prognosis [34, 35, 27]. Other applications of neural networks
abound [12, 30, 18, 13] . In this brief work we focus on a number
of problems of machine learning and pose them as optimization
problems. Hopefully this will point to further applications of
optimization to the burgeoning field of machine learning.
