Do we need to learn Optimization to build Machine Learning Models ?

Do we need optimization for Machine Learning ?

The answer is No and Yes.

If you are a beginner, just getting started, you can build a fair bit of stuff without in depth knowledge of an advanced topic such as optimization. However, if you are looking to be an expert, where you write custom models, a course on optimization helps you craft custom solutions and make better choices w.r.t to the solutions (models and solvers) available.

The above video, talks about whether and how optimization is helpful in our Machine Learning Journey.

To summarize, here are four ways a course on optimization helps us in our Machine Learning journey.

1.  Understanding “Loss Functions” that we commonly use when we build ML models.
2.  Understanding the choice of solvers for various ML models
3. Writing new Loss functions
4. Writing new solvers and updates for your custom loss function.

Hope you all have fun learning Optimization in your Machine Learning journey !