Azure ML

We talk about Machine Learning on Azure and forecast analysis with python and jupyter notebook.

Agenda

Presentation 🕥 (10:30)
- Who Is Proge-Software
- Who Am I
Introduction (10:45)
- Python
  - Pandas
  - NumPy
  - Matplotlib
  - Seaborn
  - Scikit-learn
- Jupyter Notebook
- Anaconda
- Machine Learning
  - Supervised Learning
  - Unsupervised Learning
  - Reinforcement learning
- Preprocessing data
- Forecast Analysis
  - ARIMA: Autoregressive integrated moving average
  - Prophet
Azure Machine Learning Studio 🕦 (11:30)
- What is Azure ML Studio
  - Differences with Classic version of the platform
- Why do we need it
- How does it work
  - Compute
    - Compute Instances
    - Compute Clusters
    - Inference Clusters
    - Attached Compute
  - Storage
  - Notebooks
  - Work with Computes
    - JupyterLab
    - Command Line
Demo 🕛 (12:00)
- Setup Azure ML Studio
- Data
  - Preprocessing the Data
  - Code explanation
- Spot and treat outliers
- Train a Prophet model
- Model evaluation
- Hyperparameter tuning
Where to go next (12:20)
- Artificial Neural Networks
- Deep Learning
- Frameworks for deep learning
  - Tensorflow
  - Keras
  - PyTorch
- Types of Neural Network
- NeuralProphet
Q&A (12:25)