machine learning

What is Machine Learning (ML)?

ML is derived from Artificial Intelligence (AI). It is a process of designing techniques or algorithms or models that enable a machine (any computer software) to learn from data.

There are some types of ML algorithms.

  1. Classification
    1. Supervised
    2. Unsupervised
    3. Semi-supervised
  2. Regression
    1. Linear
    2. Logistic
  3. Association Rule Mining

Who is a Machine Learning Engineer (MLE)?

An MLE is a professional who uses ML technologies to design ML models, to evaluate them, to publish them to production and then monitor alongwith version management.

Basically MLE gets more and more polished with the level of hands on experience, exposure to datasets and analytical competitiveness, which is gained via strong fundamentals of statistics, model diagnostics and mathmatical modeling.

There is wide pool of ML technologies across the industry. There are different tools for different phases of model design, development/tracking, testing and publishing.

Model Design and Development

  • Azure Databricks
    • clusters
  • Python libraries:
    • sklearn, coremltools
  • MLFlow

Model Tracking

  • Wandb.ai
  • neptune
  • tensorboard

Model Publishing and Maintenance

  • Kubernetes containers
  • Azure DevOps

ML Companies

WANDB, Neptune

Jobs in ML

Practice Platforms in ML

Kaggle

Books on ML

  1. Approaching (Almost) Any Machine Learning Problem 2020

Conferences on ML

  1. International Conference on Machine Learning
  2. MLConf
  3. Machine Learning Prague

Research Journals on ML

  1. Journal of Machine Learning Research
Acknowledgments
References