For managers and developers
Introduction to ML
Key concepts, use cases, possibilities, and limitations. How to choose the best ML solution / tools / APIs / platforms for your organization.
Decisions from predictions
Interpreting predictive models, confidence values, and errors. Using predictions to create new value in your application / business / process.
- Supervised: Decision Trees, Random Forests, Gradient Boosting, Logistic Regression
- Unsupervised: Anomaly Detection (isolation forests), Clustering (k-means)
- Deep Learning: Neural Networks, Transfer Learning
- Data processing pipelines: cleaning, natural language processing, feature engineering
- Model selection: evaluation, performance metrics, cross-validation, parameter tuning
- Deployment in production
- REST APIs for real-world ML
- Basics: Docker, Anaconda, Jupyter, Flask
- Open source ML: Scikit-learn, SKLL, Xgboost, Hyperopt, Keras, Tensorflow, Pandas
- ML platforms: Dataiku, BigML, Microsoft Azure ML