Skip to content

Latest commit

 

History

History
43 lines (28 loc) · 2.67 KB

File metadata and controls

43 lines (28 loc) · 2.67 KB

Machine Learning Keyword Mindmap

Mindmaps summarizing and categorizing Machine Learning keywords, from Data Processing to Deep Learning.

Why keywords?

Because the details can be found in the original articles/lectures. What I focus on in this project is the big picture of machine learning and thus, there is NO algorithm explanation in any provided mind maps.

Some parts in the mindmaps are blank, but I'm planning to build a complete mindmap in the future. Currently, the projects includes the following mindmaps:

  • Data Processing
  • Machine learning process
  • Machine learning algorithm
  • Convolutional neural network (comming soon)
  • Sequence model (comming soon)

Tools and Notations

I have built the mindmaps with XMind on Window environments but the software is available in macOS and Linux. The images are generated and available under folder images.

Notiations are used in all mindmaps includes:

  • capital letter (e.g., Principal component analysis) when refering to a method/algorithm
  • normal letter (e.g., dimensionality reduction) when refering to a category
  • star (e.g., algorithm selection (*)) when its details are given in other parts
  • bang! bang! (.e.g, !! ensemble models get better results but they are only for competitions (NOT productions)) for a note

1. Data Processing

Data processing, as an infomraiton processing, is to collect and manipulate data to produce meaningful information. The techniques in this mindmap is almost applied to a data type of number, special techniques for other types (e.g., speech) will be covered in other part (e.g., Sequence model part).

alt text

2. Machine learning process

In addition to the implementation/evaluation of a single model, the machine learning process should include all steps of project development where multiple models might be developed in parallel. Also, the procss covers checking points when new data arrive.

alt text

3. Machine learning algorithm

Well-known algorithms are listed and categorized into groups so that readers can remember easily. If an algorithm is used for multiple purposes, it will be listed for major purposes and may be omitted in minor one. As mentioned earlier, I am planning to complete the mindmap in the future.

alt text

4. Convolutional neural network

5. Sequence model