Machine Learning with PyTorch

In the course Machine Learning with PyTorch from SpiralTrain data scientists, data engineers and aspiring machine learning practitioners learn how to harness the power of the PyTorch framework to crea

Intro PyTorch

The course Machine Learning with PyTorch starts with an introduction to PyTorch, covering the basic principles of tensors, autograd and the PyTorch ecosystem.

Linear Regression

Subsequently linear regression in PyTorch for predicting results is discussed, including optimization with gradient descent, loss functions, regularization techniques and evaluation metrics.

Neural Networks

Then neural networks with PyTorch are treated, where activation functions, backpropagation and optimization algorithms are explained.

Classification

Classification tasks in PyTorch are also covered with logistic regression and cross entropy losses. Both binary and multi-class classification are treated.

Model Building

And model building is also on the program of the course Machine Learning with PyTorch. Here it is explained how more complex models can be based on fundamental building blocks, using feature engineering, categorical variables and hyperparameter tuning.

Natural Language Processing

Then Natural Language Processing with PyTorch is explained. The use of text classification, named entity recognition and sequence to sequence models for machine translations is covered.

Reinforcement Learning

And reinforcement learning with PyTorch is also on the program. Among others, Markov Decision Processes, Q-Learning, Policy Gradients and Actor-Critic Methods are discussed then.

Image Processing

The use of PyTorch for image processing is also covered, including classification, object detection and semantic segmentation.

Model Optimization

Finally attention is paid to optimizing machine learning models in PyTorch with the goal to improve performance and efficiency. Techniques such as batch normalization, hyperparameter tuning and pruning are treated then.

Audience Course Machine Learning with PyTorch

The course Machine Learning with PyTorch is intended for data scientists who want to use Python and the Torch machine learning library to create models and make predictions.

Prerequisites training Machine Learning with PyTorch

To participate in this course, knowledge of and experience with Python is required and knowledge of data analysis libraries such as Numpy and Pandas is desirable.

Realization course Machine Learning with PyTorch

The theory is discussed through presentations. Illustrative demos clarify the concepts. The theory is interchanged with exercises.

Certificate course Machine Learning with PyTorch

After successfully completing the course, attendants will receive a certificate of participation in Machine Learning with PyTorch.

Modules

Module 1 : Intro PyTorch

• Machine Learning Intro
• Overview of PyTorch
• Installing Anaconda
• Setting Up PyTorch
• PyTorch Tensors
• Tensor Operations
• Simple Neural Networks
• Datasets and DataLoaders
• Fundamentals of Autograd
• Model Evaluation Metrics

Module 2 : Linear Regression

• Linear Regression in PyTorch
• Gradient Descent Optimization
• Mean Squared Error
• Regularization Techniques
• Feature Scaling
• Feature Normalization
• Categorical Features
• Model Evaluation Metrics
• RMSE, MAE, R-squared
• Hyperparameter Tuning

Module 3 : Neural Networks

• Neural Networks Intro
• Building NN with PyTorch
• Multiple Layers of Arrays
• Convolutional Neural Networks
• Activation Functions
• Loss Functions
• Backpropagation
• Gradient Descent
• Stochastic Gradient Descent
• Recurrent Neural Networks

Module 4 : Classification

• Logistic Regression
• Binary Classification
• Multi-class Classification
• Cross-Entropy
• Confusion Matrix
• Precision and Recall
• ROC Curve
• Handling Imbalanced Data
• Regularization Techniques
• Hyperparameter Tuning

Module 5 : Model Building

• PyTorch Models
• Model Components
• Parameters
• Common Layer Types
• Linear Layers
• Convolutional Layers
• Input Channels
• Recurrent Layers
• Transformers
• Data Manipulation Layers

Module 6 : Natural Language Processing

• NLP Overview
• Text Preprocessing
• Tokenization
• Stopword Removal
• Spam Detection
• Bag-of-Words
• Word Embedding
• Sentiment Analysis
• Attention Mechanisms
• Transformer Models

Module 7 : Reinforcement Learning

• Intro Reinforcement Learning
• Markov Decision Processes
• Q-Learning
• Deep Q-Networks
• Policy Gradient Methods
• Actor-Critic Methods
• Proximal Policy Optimization
• Deep Policy Gradient

Module 8 : Image Processing

• Image Preprocessing
• Resizing and Normalization
• Convolution Layer
• Convolutional Neural Networks
• Object Detection
• Transfer Learning
• Semantic Segmentation
• Image Captioning

Module 9 : Model Optimization

• Profiling PyTorch
• Profiler With TensorBoard
• Hyperparameter tuning
• Parametrizations
• Pruning
• torch.compile
• Dynamic Quantization
• High-Performance Transformers