Content gap: Documentation for the Built-in algorithms is incomplete
What did you find confusing? Please describe.
I had a look at our current docs and found gaps in the Algorithm documentations between the AWS dev guide and readthedocs.io.
The AWS dev guide lists all algos that are officially released (left column below), but some of them are not documented in readthedocs.io (right column below).
| AWS dev guide | pySDK in First-Party Algorithms |
|---|---|
| BlazingText | (missing) |
| DeepAR Forecasting | (missing) |
| Factorization Machines | FactorizationMachines |
| Image Classification Algorithm | (missing) |
| IP Insights | IP Insights |
| K-Means Algorithm | K-means |
| K-Nearest Neighbors (k-NN) Algorithm | K-Nearest Neighbors |
| Latent Dirichlet Allocation (LDA) | LDA |
| Linear learner algorithm | LinearLearner |
| Neural Topic Model (NTM) Algorithm | NTM |
| Object2Vec | Object2Vec |
| Object Detection Algorithm | (missing) |
| Principal Component Analysis (PCA) Algorithm | PCA |
| Random Cut Forest (RCF) Algorithm | Random Cut Forest |
| Semantic Segmentation | (missing) |
| Sequence to Sequence (seq2seq) | (missing) |
| XGBoost | XGBoost (is under the Framework node) |
In case of XGBoost, I suggest to list it with the other algos and provide a link to the exisiting XGBoost in the Framework node with a short explanation of why XGBoost is under Frameworks.
Describe how documentation can be improved
- Add the 6 missing algorithms to the readthedoc.io 1p algorithm section.
- Add XGBoost to the 1p algorithm section as well, and link to the existing XGBoost documentation under the Frameworks node.