Python for machine learning - course RUB 49,990. from Specialist, training, Date: November 27, 2023.

The course focuses on the use of Python for machine learning technologies. The main libraries used in machine learning and methods for using them are studied.
The course is taught based on the machine learning libraries Scikit-Learn and TensorFlow. During the learning process, development tools Jupyter Notebook and Jupyter-Lab are actively used. Google Colab cloud computing resources are also used.
The course is recommended for Python programmers and data science specialists.

You will learn:
apply machine learning technologies;
use classification, regression, clustering models;
use data preprocessing methods and machine learning covers;
select machine learning models based on the quality of their work;
deepen your knowledge of Python, master modern tools for developing machine learning models.

Module 1. Statement of the machine learning problem (2 ac. h.)
Three types of machine learning.
Scheme for constructing machine learning systems.
Necessary tools.
Practice creating a working environment and using Anaconda and Jupyter Notebook.

Module 2. Training with a teacher. Classification problem (6 ak. h.)
The concept of a neural network.
Perceptron. Definition, implementation and training.
Adaptive linear neuron. Definition, implementation and training.
Gradient descent method.
Stochastic gradient descent in an adaptive linear neuron.
Dynamic learning on big data.
Practice with model and Scikit-Learn datasets.
Practice in Jupyter Notebook.
Module 3. The scikit-learn library, its main features (6 ac. h.)
Perceptron training.
Logistic regression method.
Support vector machine.
The k-nearest neighbors method.
Decision trees.
An ensemble of arbitrary trees.
Practice in Jupyter Notebook.
Module 4. Training with a teacher. Regression problem (6 ak. h.)
Training with a teacher. Regression problem.
Linear regression.
Least square method.
Gradient descent method.
Working with emissions. RANSAC() algorithm.
Bootstrap technique for regression coefficients.
Assessing the quality of the regression model.
Basics of nonlinear regression.
Practice in Jupyter Notebook.
Module 5. Data preprocessing, selection of features and models (6 ac. h.)
Handling missing data.
Processing categories.
Preprocessing and training models.
Selection of significant features.
Exploring significant features using random forests.
Machine learning pipelines.
Cross-validation and model selection.
Practice in Jupyter Notebook.
Module 6. Learning without a teacher. Cluster analysis (6 ac. h.)
K-means method.
Principal Component Analysis.
Linear discriminant analysis.
Hierarchical clustering.
Clustering by density.
Practice in Jupyter Notebook.
Module 7. Fundamentals of deep learning (8 ac. h.)
Graphs and distributed computing.
Installing Tensorflow.
Tensors.
Data sets.
Application interface (API) Keras.
Linear regression with stochastic gradient descent.
Classification by two-layer neural network of Iris data.
Activation functions of neuron layers.
MNIST data classification.
Practice in Google Colab.