RailsConf 2023 - Forecasting the Future: An Introduction to Machine Learning for... by Landon Gray

Learn how to apply machine learning in Ruby, a native language for many developers. Discover linear regression, data preparation, and existing libraries like Daru and Numo to simplify tasks. No Ph.D. required!

Key takeaways
  • Machine learning in Ruby, not just Python: The speaker highlights the need for machine learning in Ruby development, as many developers prefer to work with their native language.
  • Linear regression: The speaker focuses on linear regression, a simple and easy-to-understand machine learning algorithm, for predicting weather data.
  • Data preparation is key: The speaker emphasizes that 80% of the time is spent on data preparation, including cleaning, processing, and handling missing values.
  • Use existing libraries: The speaker recommends using existing libraries like Daru, Remale, and Numo to simplify data manipulation and machine learning tasks.
  • Keep it simple: The speaker suggests starting with a simple project and scaling up as needed, rather than trying to tackle complex machine learning tasks from the beginning.
  • RubyConf: The speaker will publish the project on GitHub and encourage attendees to download and tinker with the code.
  • No Ph.D. required: The speaker believes that machine learning is accessible to developers without a Ph.D. in data science.
  • Validation is crucial: The speaker stresses the importance of validating machine learning models against testing data to ensure accuracy and reliability.
  • Ruby gems are available: The speaker mentions that there are Ruby gems available for machine learning, such as Numo and Daru, which make data manipulation and processing easier.
  • Imputing missing values: The speaker suggests imputing missing values using averages or other methods to avoid duplicates and maintain data accuracy.
  • Focusing on Ruby: The speaker encourages developers to focus on Ruby and not feel pressured to learn multiple programming languages.