Course objectives
To get familiarized with Python as a tool for automated data processing, which involves data analysis and visualization, model building, assessment and validation, prediction, as well as reporting results of the analysis.
Training scope
- Introduction to statistical data analysis
- Supervised and unsupervised learning
- Classes and examples of models
- Methods of model estimation/calibration
- Principles of statistical hypothesis testing
- Model validation and selection
- Overfitting (bias-variance trade-off)
- Data Science techniques that will be implemented:
- Linear and nonlinear regression models
- Artificial neural networks
- Decision trees and random forests
- Bayesian classifiers
- Ensemble modelling
- Confusion matrix, ROC curve
- Model quality statistics
- Cross-validation
- Model development with Python:
- Python as a programming language
- PyCharm – an IDE for Python
- Jupyter Notebooks – documents with live code
- Pandas for:
- Import and export of data
- Data structures and data manipulation
- NumPy for linear algebra and numerical computation
- Scikit-learn for:
- Regression
- Classification
- Clustering
- Model validation and selection
Audience
Data Scientists or anyone who wants to use Python as a tool for automated data analysis. This is an introductory level training. Utilized ML methods and programming techniques will be explained, but it is assumed that participants have a basic experience with Python.