Alex is a scientist (Deep Learning, Machine Learning, Computer Vision, Artificial Intelligence), speaker, and engineering executive with over 15 years of intl. experience in research (Columbia U., KAIST) and product impact at scale (DigitalOcean, Yahoo, Telefónica, IDIAP-EPFL, Fuji Xerox, IBM, Siemens, and AT&T Bell Labs) in the USA, Japan, Chile, Switzerland, Spain, and South Korea. He has authored over 100 patents and publications (h-index 35), has a Ph.D. from Columbia University and is a frequent invited speaker at an international conferences and industry events.
There are many opportunities in applying Machine Learning, whether as an individual developer or in a business. But how do you get started? In this talk, I will first clearly give an overview that separates fact from fiction, and propose processes to find opportunities for applying ML. This will include understanding where ML can have the biggest impact while avoiding common pitfalls. I will emphasize how sometimes, improvements in processes can significantly outweigh algorithmic improvements. We will examine, for example, data collection and quality, definitions used (e.g, for labeling), metrics, objective functions, overfitting, and the cost of different types of errors, among others. The perspectives presented will help attendees get a better grasp of applying ML in real-world scenarios, including the selection of specific algorithms for different tasks. Key takeaways include:
- How to identify data sources and data quality issues
- How to come up with the right metrics
- How to deal with different types of errors and understand their impact
- How to examine and improve processes that impact the application of ML