Goals for 2021

Published on Wednesday, 09 December, 2020

Overarching Goals

• Provide an ML solution that provides value. Value can be either a useful open-source project or a profitable solution.
• Discover a novel ML approach worth publishing. Publish it.

Site Goals

• Provide specific weekly status updates toward my objectives
• Write a reader-focused post twice monthly
• Make this list :)

Technical Goals

• Complete Gilbert Strangs Introduction to Linear Algebra
• Implement fundamental Reinforcement Learning algorithms [refinement needed]
• Implement SOTA Deep RL algorithms [e.g. PPO; refinement needed]
• Apply learned technical expertise to a monthly unique or novel problem (e.g. dont simply use existing OpenAI gyms for everything. Possibly develop new environments)

---

Objectives

• Establish a commitment to progress
• Be less compulsive in my work habits by creating a steady stream of content

Technical Expertise

• Review of ML fundamentals, including linear algebra, calculus, and statistics
• Mastery of Reinforcement Learning concepts

Establish Context

• Create processes for developing technical expertise in context
• applying a method not just to a problem, but a set of problems
• Identifying and understanding alternatives

Collaborate

• Develop content and projects with the assistance of others
• Contribute to projects other than my own
• Have fun along the way

---

Audience (in this order)

• Me
• ML students and professionals, particularly those working at the crossroads of Reinforcement Learning and NLP
• Ancillary: professional contacts (e.g. future employers)