Python - Backpropagation in neural networks

• Knowledge of Python programming at the PYTHON_INTRO course level, but knowledge at the PYTHON_ADV course level is an advantage
• Knowledge of the basics of data analysis at the level of the PYTHON_DATAAN course
• Knowledge of the basics of machine learning at the level of the PYTHON_ML_INTRO course
• Knowledge of the basics of neural networks at the level of the PYTHON_ML_NN course
• Knowledge of the basics of Backpropagation in neural networks at the level of the PYTHON_ML_CNN course

• Professional explanation with practical examples, exercises on computers.

• Presentation of the subject matter in printed or online form.

Day 1:

• Introduction to machine learning and neural networks
• Forward propagation in a neural network
• Activation function in a neural network
• Design of a simple neural network

Day 2:

• Backpropagation of the error in the neural network
• Gradient descent and its variants
• Calculation of gradients for each network parameter
• Practical exercises for backpropagation implementation

Day 3:

• Multilayer neural networks
• Network training using backpropagation
• Solving the problem of overtraining
• Practical exercises for training multilayer networks

Day 4:

• Convolutional neural networks and their foundations
• Pooling, convolution and strided convolution
• Design and training of a simple convolutional network
• Practical exercise on convolutional networks

Day 5:

• Recurrent neural networks and their foundations
• LSTM and GRU networks
• Design and training of recurrent network
• Evaluation of results and possibilities for improvement
• Practical exercise on recurrent networks

Previous courses

Python I - introduction to programming [PYTHON_INTRO]

Python - data analysis (Pandas) [PYTHON_DATAN]

Python - Neural networks [PYTHON_ML_NN]

Python - Convolutional neural networks [PYTHON_ML_CNN]

Next courses

Python - Deep Learning [PYTHON_ML_DP]