A step-by-step process for approaching ML problems in Real life:
- Understanding the Business Requirements and the Nature of the Available Data.
- Classify the problem as Supervised/Unsupervised as well as Regression/Classification (in advance).
- Download, Clean & Explore data and Create New Features (if required) that may improve models.
- Create Training/Validation/Test sets of data and prepare the data for training ML models.
- Create a quick & easy baseline model to evaluate and benchmark future models.
- Select a modeling strategy, train a model, and tune hyperparameters to achieve optimal fit.
- Experiment and combine results from multiple strategies to get a better result.
- Interpret models, study individual predictions, and present your findings to the stakeholders.
Check out the full scoop here: ML Cheat Codes 2024
| # | Topics | Links |
|---|---|---|
| 1 | Data-Gathering |  |
| 2 | Understanding-Data | |
| 3 | Feature-Engineering | |
| 4 | Outlier Handlings | |
| 5 | PCA (Principal Component Analysis) | |
| 6 | UMAP (Uniform Manifold Approximation and Projection) | |
| 7 | Tensors | |
| 8 | Linear-Regression | |
| 9 | Logistic-Regression | |
| 10 | DT-and-RF | |
| 11 | Voting-Ensemble | |
| 12 | XGBoost | |
| 13 | Support-Vector-Machine | |
| 14 | Unsupervised-Clustering | |
This is the Ultimate Cheat Code for Machine Learning is what you will need in 2024. If you are beignner in Data Science or even if you're a seasoned expert these codes will be your quick syntax reference.