course-zoomcamp

Machine Learning Zoomcamp

Course overview video and slides

Register for the course here

Public calendar (subscribing works from desktop only)

1. Introduction to Machine Learning

• 1.1 Introduction to Machine Learning
• 1.2 ML vs Rule-Based Systems
• 1.3 Supervised Machine Learning
• 1.4 CRISP-DM
• 1.5 Model Selection Process
• 1.6 Setting up the Environment
• 1.7 Introduction to NumPy
• 1.8 Linear Algebra Refresher
• 1.9 Introduction to Pandas
• 1.10 Summary
• 1.11 Homework

2. Machine Learning for Regression

• 2.1 Car price prediction project
• 2.2 EDA
• 2.3 Setting up the validation framework
• 2.4 Linear regression
• 2.5 Applying linear regression
• 2.6 Linear regression: matrix form
• 2.7 Training linear regression: Normal equation
• 2.8 Implementing linear regression with NumPy
• 2.9 Car price - the baseline solution
• 2.10 Evaluating the model with RMSE
• 2.11 Validating the model
• 2.12 Feature engineering
• 2.13 Categorical variables
• 2.14 Regularization
• 2.15 Tuning the model
• 2.16 Using the model
• 2.17 Car price prediction project summary
• 2.18 explore more
• 2.19 homework

3. Machine Learning for Classification

• Churn prediction project
• Initial data preparation
• Setting up the validation framework
• EDA
• Feature importance: Churn rate
• Feature importance: Risk ratio
• Feature importance: Mutual information
• Feature importance: Correlation
• One-hot Encoding
• Logistic regression
• Logistic regression with sklearn
• Using logistic regression (?)
• Logistic regression - interpretation
• Applying logistic regression
• Summary
• Explore more
• Homework

4. Evaluation Metrics for Classification

• Evaluation metrics
• Accuracy
• Baseline solution
• Confusion table
• Calculating the confusion table
• Precision and recall
• Receiver operating characteristic (ROC)
• ROC: Random baseline
• ROC: Ideal model
• The ROC Curve
• ROC curve with skilearn
• Area under the ROC Curve (AUC)
• intepretation of AUC
• K-fold cross-validation
• Selecting the best parameter C
• summary
• explore more
• homework

5. Deploying Machine Learning Models

• Using the model
• pickle
• Deploying a model as a Web Service
• Introduction to Flask
• Model serving with flask
• Managing dependencies with Pipenv
• Introduction to Docker
• Testing it locally
• AWS beanstalk
• summary
• explore more
• homework

6. Decision Trees and Ensemble Learning

• Credit risk scoring project
• Data cleaning
• Data preparation
• Decision trees
• Decision tree learning algorithm
• impurity
• split
• stopping criteria
• Decision trees parameter tuning
• Ensembles and random forest
• Random forest in sklearn
• Random forest parameter tuning
• Gradient boosting
• eXtreme Gradient Boosting - XGBoost
• training
• watchlist
• XGBoost parameter tuning
• learning rate
• max_depth
• min_child_weight
• Testing the final model
• summary
• explore more
• homework

7. Midterm Project

8. Neural Networks and Deep Learning

• Clothes classification project
• TensorFlow and Keras
• loading the images
• etc
• Using a pre-trained model
• CNNs: convolutional layers
• CNNs: dense layers
• Transfer learning
• Creating the clothes classification model
• Keras functional components
• optimizer
• training the model
• Learning Rate
• Model checkpointing
• Adding more layers
• Dropout
• Data augmentation
• Training a larger clothes classification model
• Using the model with Keras
• summary
• explore more
• homework

9. Serverless Deep Learning

• intro
• serverless and AWS Lambda
• tensorflow-lite
• converting the model to TF-lite
• preparing images
• using the model in TF-lite
• putting everything together in a Lambda function
• preparing the docker image
  • testing the image locally
• pusting the image to ECR
• creating the lambda function
• creating the API gateway
• summary
• explore more
• homework

10. Kubernetes and TensorFlow-Serving

• intro, serving architecture overview
• saved_model format
• tensorflow-serving
  • running TF-serving locally
• communicating with tf-serving from Jupyter
• creating the gateway service
• introduction to Kubernetes
• creating a cluster on AWS (article)
• preparing the images
  • the TF-serving image
  • the gateway image
• deploying to Kubernetes
  • deployment for tf-serving
  • service for tf-serving
• creating the gateway on Kubernetes
  • deploymnet
  • servince - load balancer
• testing it
• deleting the cluster
• summary
• explore more
• homework

11. Kubeflow and KFServing

• intro
• installing Kubeflow on AWS
• preparing the model: uploading to S3
• deploying TF models with KF-serving
• accessing the model
• tranformers
• testing it
• deleting the cluster
• summary
• explore more
• homework - no homework

12. Capstone Project

13. Article